b, a.q-1
P53 AND K¡-67 (MlB-l) EXPRESSION
IN
ODONTOGENIC KERATOCYSTS
Thesis submitted in partial fulfillment of the requirements for the
degree of Master of Dental SurgerY
Adi Rahmadsyah DDS (Medan)
March 1998
Faculty of Dentistry
The University of Adelaide
I l This thesis is dedicated to
Ayah & lbu
Hanim, Sasha and Della TABLE OF CONTENTS
Page
Declaration
Acknowledgements
Chapter 1 lntroduction 1
Ghapter 2 Literature Review 4
2.1 The odontogenic keratocYst 5
2.1.1 Clinical features 6
2.1.2 Radiologicalfeatures 11
2.1.3 Histopathological features 12
2.1.4 Behaviour 16
2.1.5 Histogenesis & pathogenesis of odontogenic 18
keratocysts
2.2 P53 20
2.2.1 lnitial identification of P53 20
2.2.2 P53 structural characteristics 21
2.2.3 Molecular weight and location of p53 22
2.2.4 p53 protein level in cells 22
2.2.5 Types of p53 23
2.2.5.1 Wild-type p53 23
2.2.5.2 Mutant p53 24
2.2.6 Biological function of P53 27
2.2.6.1 Checkpoint of cell cYcle 28 2.2.6.2 Apoptosis 29
2.2.6.3 Cellular differentiation 29
2.2.7 Detection of p53 in Cells 30
2.2.8 Application of p53 in tumour diagnosis 30
2.2.9 P53 protein expression in oral lesion 33
2.3 Ki-67 and its functions 35
2.3.1 Ki-67 35
2.3.2 MIB-1 as an alternative to Ki-67 37
2.3.3 Ki-67lMlB-1 as indicators of cellular 39
proliferation
2.3.4 Expression of Ki-67 / MIB-1 in tumour cells 39
Chapter 3. Materials and Methods 43
3.1 Materials 44
3.1.1 Case selection 44
3.2 Methods 46
3.2.1 Preparation of histologic material 46
3.2.2. Assessment of clinical data and histological 47
material
3.2.2.1 Clinical data 47
3.2.2.1.1 Age 47
3.2.2.1.2 Sex 47
3.2.2.1.3 Site 47
3.2.2.2 Hlstological assessmerlI 48
3.2.2.2.1 Epithelial lining 48 3.2.2.2.2 Number of cysts and behaviour 48
3.2.2.2.3 lnflammation 48
3.2.4 lmmunohistochemical phase 49
Chapter 4 Results 50
4.1 Clinical data 51
4.1.1 Age distribution of total OK Pool 51
4.1.2 Age distribution of selected OK cases 51
4.1.3. Sex distribution of total OK pool subjects and 54
immunohistochemistry pool subjects
4.1.4 Site distribution of total OK Pool 54
4.2. Histologic obseruations 58
4.2.1 Epithelial lining 58
4.2.2 I mmunohistochemical staining 62
4.2.2.1 P53 staining 62
4.2.2.2 Relationship between p53 staining and 66
4.2.2.3 MIB-1 staining 68
4.2.2.4 Relationship between p53 staining and MIB-1 70
positivity and inflammation
4.2.2.5 Relationship between p53 staining and MIB-1 72
staining
4.2.2.6 Ameloblastoma 74 Chapter 5 Discussion 77
5.1 Clinical and histological findings 78
5.2 Technical aspects of immunostaining 79
5.3 Pre-treatment effects on staining 81
5.4 p53 staining 82
5.5 MIB-1 staining 83
5.6 P53 staining and inflammation 83
5.7 MIB-1 staining and inflammation 84
5.8 Ameloblastoma 84
Conclusions 86
Bibliography 88 DECLARATION
This work contains no material which has been accepted for the award of any other degree of diptoma in any university or tertiary institution and to the best of my knowledge and belief contains no material previously published or written by another person, except where due reference has been made in the text.
I give consent to this copy of my thesis being made available for loan and photocopying
Adi R h March 1998 ACKNOWLEDGEMENTS
I am deeply grateful for the patience and support of Associate Professor David
F. Wilson, the supervisor of this project. He has donated extraordinary
amounts of time, and has provided much invaluable advice on aspects of
scientific research and methodology.
Many thanks also go to Dr Angela Pierce for her assistance and advice, and
to Sandie Powell and Margaret Leppard for laboratory and photography
assistance.
Thanks especially to the Director and Staff of the lnstitute of Medical and
Veterinary Science (IMVS), Adelaide for allowing access to laboratories as a r+ rrJ part of my training. Thanks also to Peter Gilan, IMVS for helping with the
immunohistochemical staining of the specimens. A thankyou also to
Professor Grant Townsend for his assistance in statistical analysis.
Sincere thanks are also expressed to the Australian Agency for lnternational
Development (AusAlD) for sponsoring me in this Master of Dental Surgery
program.
My thanks are extended to Ms Jo Seton, ACUE, University of Adelaide for
invaluable assistance in editing draft materials. Special thanks also to Mrs.
Juliefllugo for her help in the final preparation of this thesis.
þ CHAPTER 1
INTRODUCTION
1 Odontogenic keratocysts (OKs) are developmental cysts of the jaws and are
known to express growth and behavioural characteristics more analogous to
benign neoplasms than to simple jaw cysts. A number of studies on
odontogenic keratocysts have emphasized that the aggressive behavior of
these cysts might be associated with epithelial cyst lining activities'
:. i The unusual behavior and the characteristic histological features of OKs have
been the subject of many studies. More recently the results of a number of
immunohistochemical studies have been reported in the literature. For
example a small number of studies have indicated that p53 (p53 protein is a
product of the p53 gene and acts as a checkpoint which monitors the efficacy
and completion of the obligatory events in the cell cycle and blocks
uncontrolled cell division) expression in OKs is higher than in other jaw cysts
and that the location of p53 positive cells was more likely in the suprabasal
epithelial cells. However, in these studies it is apparent that there is
considerable variability when the results of different investigations are
compared.
ln addition to studies concerning p53 distribution in OKs, there have been a
very small number of studies investigating the presence and distribution of
anothercell proliferation marker Ki-67 (MlB-1). To date the results of these
few studies have indicated that OKs express Ki-67 (MlB-1) at a level greater
than for p53.
2 Apart from recent studies by Slootweg (1995) who studied the distribution of p53 and Ki-67 in a sample of 13 OKs and Li et al (1996) who undertook a similar study on a sample of 22 OKs there are few data available describing the correlated distribution of p53 and Ki-67 (MlB-1) in OKs.
The aims of the present study were to:
. Analyse p53 and MIB-1 distribution in a large sample of oKs.
. Determine whether or not staining patterns correspond for each
marker.
A small sample (9) of ameloblastomas were also analysed'
Fifty cases of OKs were examined for p53 and MIB-1 (Ki-67) expression using immunohistochemical techniques following microwave pretreatment of histological sections. The primary antibody used for immunohistochemical staining of OKs and ameloblastoma were monoclonal p53 (Dako-DO-7) and
MIB-1 (Zymed).
J CHAPTER 2
LITERATURE REVIEW
4 2.1. THE ODONTOGENIC KERATOCYST
Odontogenic keratocysts (OKs) are developmental cysts of the jaws. The cysts are thought to originate from odontogenic epithelium, such as dental lamina or its remnants within bone or oral mucosa, and from the proliferation of basal cells of the oral epithelium (Wilson and Shear,1992i Browne, 1970b).
Histologically, these cysts are characterised by keratin formation.
The term odontogenic keratocyst was introduced by Philipsen (1956) and its essential clinical and radiological features were described by Pindborg and
Hansen (1963). At first, the term was applied to any odontogenic cyst which had undergone keratinization. However, Hansen (1967) concluded that odontogenic keratocysts were not of inflammatory origin.
Prior to classification of their terminology in the nineteen sixties OKs had long been known by different terms. ln his book Cysts of the Oral Region, Shear
(1992) stated that this cyst was described as a dermoid cyst by Mikulicz
(1876) and then as a cholesteatoma by Hauer (1926) and Kostecka (1929)'
Although some authors had suggested the term "primordial cyst" as a synonymous term for odontogenic keratocyst, the former term is still not accepted worldwide because of the lack of evidence to support it'
The term primordial cyst had been used to describe cysts developing in place
of a tooth through degeneration of the enarnel organJhsterm was first used by Robinson (1945) and, subsequently, used by the World Health
5 Organization in 1971. ln 1985, Main and Shear suggested the term
"odontogenic keratocyst" be used to designate this cyst of the jaws'
Histologically, the OK has a characteristic appearance. lt can occur as a single cyst or multiple cysts. lt can also be associated with the Basal Cell
Naevus Syndrome (BCNS).
2.1.1. Glinical features
The clinical features of OKs, such aS age, incidence, Sex, the site of occurrence and the number and size of lesions are relatively variable.
The incidence of OKs reported in studies conducted between the 1960s and
1990s also varies. Compared to other jaw cysts, the incidence of OKs is relatively low. A study by Toller (1967) reported that 33 (11.0%) of 300 jaw cysts of all types were OKs. ln contrast, a study by Hoffmeister and Harle
(1985) showed that only 51 (1 .5 %) of 3353 cases of jaw cysts recorded were
OKs.
The incidence of OKs varies between white and black populations, with the
white population more likely to suffer OKs than the black population,
According to The World Standard Population, the incidence of OKs is higher in white males than in white females, black males and black females
(Rachanis and Shear 1978).
6 OKs have been found in patients ranging in age between the first and ninth decades. However, the highest incidence of OKs (40-60 %) occurs between the second and third decades. Some authors have reported a bimodal age distribution with a frequency peak occurring in younger patients (second and third decades) and a second peak in the fifth decade or later (Toller 1967;
Forsell 1980; Ahlfors et al 1984; Woolgar et al 1987). Although the number of cases encountered in a younger age group studied by Rachanis and Shear
(1978) was higher than that of an older group, the latter group was actually at greater risk of diagnosis of OKs. lt is unlikely that these cases represent different varieties of OKs. Rather it likely that the cysts in the older age group have been present, but undiagnosed, for many years (Brown 1975). This accords with a characteristic of OKs, namely, that they may significantly extend in a proximal - distal direction, with little or no bone expansion.
On the other hand, the average age of patients with multiple keratocysts, whether accompanied or not by Basal Cell Naevus Syndrome (BCNS), as pointed out by Woolgar et al (1987), is considerably lower than the mean age of patients with single non-recurrent keratocysts. ln this study, the average age of patients at the time of the removal of the cysts was26.2 years, with a f requency peak at the age range of 10 - 19 years. By contrast, in the study conducted by Kakarantza-Angelopoulou and Nicolatou (1990), most patients, of whom only 7 were BCNS related, were in their fifth and sixth decades.
occur more nma an tn
predominance is more prominent in blacks than in whites, the ratios being
1 1.5:1 and 5.0:1, respectively (Shear 1992). Overall, data from different
studies show a varying male to female ratio ranging from 1.35:1 (Brannon
1976) to 1:2 (Shear 1992). ln the Kakaranlza-Angelopoulou and Nicolatou
(1990) sample, however, more hhan73"/" of patients were males. ln contrast,
in patients with multiple keratocysts, with or without BCNS, females seemed
to be more affected than males. Brannon (1976) reported that the incidence
of OKs associated with BCNS was lower in males than female patients, with a
ratio 1 :1 .71 . Browne (1977) found the female to male ratio was 2:1 , whereas,
according to Woolgar et al (1987), it was 1:1.22 compared with 1:0.62 in
patients with single OKs. The reasons for the differences in these findings
have not been elucidated.
Basal Cell Naevus Syndrome, which is better known as Gorlin-Goltz
Syndrome, has as its major features multiple basal cell carcinomas, multiple
jaw cysts, skeletal anomalies, parietal and temporal bossing, calcification of
falx cerebri, palmar and plantar pits and hypogonadism. There is a repoft
describing squamous cell carcinoma arising from an odontogenic keratocyst in
a patient with this syndrome (Moos and Rennie 1986).
Although OKs can occur anywhere in the jaw, the mandible is affected more
frequently than the maxilla. The frequency of mandibular involvement in
studies from the late 1960's to the late 1980's is such that more than two
o/", thirds of the OKs were reported f rom the lower jaws (75 7", Shear 1991 ; 77
' Hansen 1967; Ito/tBrowne 19704165 o/tBÍannolT1976'J? u/o, Vedtofte and
Praetorius 1979;78 "/", Forsell 1980; 75 "/", Ahlfors et al 1984; and 69 "/",
8 Voorsmit 1981). These findings are supported by recent studies, such as that of Kakarantza-Angelopoulou and Nicolatou (1990), in which 650/" ol OK cases were found in the mandible, and the study of Brondum and Jensen (1991) which reported lhal44 (82%) of 54 cases of OKs were in the mandible.
Most of the cases found in the mandible occurred within the posterior portion of the body, with or without extension to the ascending ramus, This finding was also noted by Kakarantza-Angelopoulou and Nicolatou (1990) and by
Brondum and Jensen (1991), who reported that 50 o/" of cases involved the angle of the mandible and ascending ramus. Only 10 (18 "/") of cases were found by Brondum and Jensen (1991) in the maxilla, of which 70"/" were located in the anterior and canine region, and 30% were in the molar area.
However, Woolgar et al (1987), as cited by Shear (1992), repofied that after the age of 50 years, there is a greater possibility of OKs being found in the maxilla.
Forsell et al (1980) observed that OKs found in the maxilla, were generally smaller and seemed to be clinically detected much earlier than those in the mandible. lt was suggested that cysts in the maxilla are likely to become infected even when they are small in size.
OKs vary considerably in size. ln Forsell's (1980) study, approximately 50 %
of cases were 40 mm or more in size, cysts in the angle of the mandible and
ascending ramus in particular being in this category, while cysts in the maxilla,
and those involving the body of the mandible, were more likely to be smaller in
9 size. ln addition, the study of Brondum and Jensen (1991) showed that 4 of
44 cases had cysts of more than 50 mm in diameter, 3 cases had cysts of 20 to 49 mm and only one case had a cyst of 15 mm in diameter.
There are no characteristic features by which a clinician might be guided to the diagnosis of OKs, and patients vary widely in their clinical presentation.
On many occasions, patients are free of symptoms until the cyst has reached a large size, This is because, as previously mentioned, OKs tend to extend along the cancellous bone without causing significant bone expansion on the cortical plate until late in the clinical course of the lesion (Browne 197Oa;
Shear 1gg2). On presentation some patients will complain of pain, while those with large lesions will show evidence of bone expansion in the involved
area. About one third of maxillary keratocysts will show buccal expansion, whereas palatal expansion is rarely seen. In the mandible, about half of all cases, particularly those involving third molars and/or an ascending ramus region, will exhibit buccal expansion, but lingual expansion is also not infrequently encountered. Like many other intrabony lesions, a large keratocyst may cause displacement of teeth, proptosis of eyeballs, bone
perforation, destruction of the floor of the orbit and neurological symptoms
(Voorsmit 1981; Lund 1985; Browne 1970a; Brannon 1976). lf the cyst is
infected or inflamed it may give rise to clinical presentations including cellulitis,
abscess, trismus and fistula formation (Browne 1970a; Brannon 1976).
Most patients with OKs present with a single lesion, although it is sometimes a
very extensive one. Some patients, however, may present with multiple OKs
10 at the time of diagnosis, or may develop multiple cysts throughout their lives.
These patients may either have multiple cystic lesions of the jaws alone or the cysts may be related to the Basal Cell Naevus Syndrome (Gorlin-Goltz syndrome). This autosomal dominant condition is an inherited abnormality which expresses skin lesions, jaw and other skeletal lesions, central nervous system and eye lesions, as well as some fairly typical facial features, of which enlargement of the size of the skull, due to bossing of the frontal and parietal bones, and over-development of the supra-orbital ridges, are particularly common and noticeable features. The presence of OKs is one of the most common features of Gorlin-Goltz syndrome, affecting 65-75 "/" of all cases, and this tendency increases with age (Browne 1991; Shear 1992). No clear relationship between the occurrence of BCNS and OKs has been established, but the cysts are thought to be derived directly from dental lamina (Shear
1ee2).
2.1 .2. Radiological features
The radiographic appearance of OKs is variable. Generally it is described as a well-demarcated, round to ovoid, radiolucent area with a distinct radiopaque border. The margin can sometimes be scalloped, with a smooth periphery, or
can be partly diffuse (Browne 1970a; Brannon 1976; Shear 1992)'
OKs can also present unilocular and multilocular radiographic features. The
majority of lesions are unilocular radiolucencies, as shown in the study
conducted by Browne (1970a). ln 83 cases of OKs from both jaws, 47 (57 "/")
11 of the cases in the study were unilocular, 17 cases (20 "/") were loculated and
19 (22.O %) were multilocular lesions. This distribution is similar to the work reported by Brannon (1976) in which 32 (61 .5 "/") ol 52 cases were unilocular,
S cases (9.6 %) were bilocular, 12 (23.O %) were multilocular, and 3 had an ill- defined border,
2.1 .3. Histopathological features
The histological features of OKs are characteristically distinguishable from other types of odontogenic cysts. The histological features of an OK can be divided into three parts. These are as follows: the cyst cavity which consists of protein fluid or a lumen full of keratin or blood; the lining epithelium, with keratinised stratified squamous epithelium; and the capsule, which comprises fibrous connective tissue.
OKs, as Browne (1970b), Brannon (1977), Forsell (1980) and Shear (1992) have described, are cysts with the characteristic microscopical feature of a thin-walled, collapsed and folded-lining epithelium.
Pindborg et al (1960) (as referred to by Brannon (1977)) designated the
following histological criteria for OKs:
1. the cysts are usually lined by a very thin epithelium and uniform in
thickness, with little or no evidence of rete ridges;
2. the basal layer is well defined and is cuboidal or columnar in shape with a
palisaded arrangement;
12 3. the spinous cell layer is thin and often shows a direct transition from the
basal cell layer, and the cells commonly show intracellular edema;
4. the keratin layer can be parakeratinised and orthokeratinised and the
sudace is often corrugated.
The terms orthokeratinised and parakeratinised describe the stage of keratinisation of the epithelial layers. Parakeratinisation is described as incomplete keratinisation that demonstrates the appearance of nuclei in the keratinised layer of the epithelium, whereas ofthokeratinisation describes a complete keratinisation of the epithelial layers without nuclei in the keratinised layers (Lucas 1984; Domenici-Lombardo et al 1996).
The majority of the reported cases of OKs (80-90%) are parakeratotic, and only a few (1O-20o/") are orthokeratotic (Brannon 1977; Cohen and Shear
1980; Voorsmit 1984). Sometimes, different parts of the same cyst can be both parakeratotic and orthokeratotic. The different types of keratinisation present could possibly be due to the type of underlying basal cell layer.
A common feature of the lining epithelium of OKs is a keratinised stratified squamous epithelium 5-8 cell layers thick without rete ridges. However, there
may be acanthotic irregularities and rete ridge proliferation which can occur in
both parakeratinised and orthokeratinised cysts. The form of the rete ridges
can be irregular in shape and formation, forming, for example, a broad, blunt
type or a budding type (Brannon 1977).
13 Cysts with orthokeratinised lining often present cuboidal basal cells. On the other hand, a corrugated surface is often seen in parakeratinised cysts with a palisaded basal layer containing columnar or cuboidal cells or both.
Most of the cysts described in the study by Kakarantza-Angelopoulou and
Nicolatou (1990) were lined by a thin, stratified squamous, parakeratinised epithelium with a corrugated surface without rete pegs. The epithelia had varying thicknesses, ranging between 5 to 15 cell layers thick' Only 24'14"/" of cases had orthokeratinised epithelium and a granular cell layer. Columnar and cuboidal basal cells were found separately in some cysts or formed a mixed population in other lesions. Thirty-nine cases contained a flattened basal cell layer, which was thought to be related to the presence of orthokeratinised lining epithelium.
Some of the keratocysts described by Kakarantza-Angeloupuolo and
Nicolatou (1990) and by Brondum and Jensen (1991) had a loose connective tissue capsule with subepithelial hyalinisation, with the epithelium separated from underlying connective tissue.
According to Forsell (from Brondum and Jensen 1991) OKs can be classified
into five groups as follows;
la thin, band-like parakeratotic cyst epithelium; basal cells cuboidal or
columnar, accentuated, and palisaded;
lb epithelium composed mainly of basal cell layer only; basal cells
accentuated;
t4 orthokeratotic cyst epithelium with distinct stratum granulosum;
accentuation of basal cells not pronounced; ilt cyst epithelium exhibiting areas of orthokeratinisation and
nonkeratinisation;
IV parakeratotic cyst epithelium resembling the oral mucous membrane;
V cyst epithelium thin and parakeratotic; basal cells not accentuated.
The study by Brondum and Jensen (1991) demonstrated Thal 27 of 44 patients in their study fell into group la of the Forsell classification; no patients were in group lb; 5 were in group ll; only 1 female patient was in group lll; 3 patients were in group lV and group V contained I patients. ln their investigation, all patients had a variety of risks of recurrence according to the group into which they fell.
Toller (1967) described that cyst fluids from keratinising cysts had less than
3.5g per 1O0ml soluble protein, whereas non-keratinising cysts contain
between 5.0 - 11.0g per 100m1, with a mean of 7.09 per 100m1. Toller (1967)
concluded that OKs could be diagnosed if the cyst fluid contained less than
4.0 g per 100 ml of soluble protein.
There are some rare variations in the epithelial lining of OKS, which may be
found. Brannon (1977) and Browne (1970b) reported that the epithelial cells
may undergo metaplastic changes, as evidenced by the presence of mucous
cells (reported in 2 - 4% of cases) and ciliated cells (occurring in 1 - 4"/" of
cases). Hyaline bodies can occur, along with foci of inflammation where the
l5 characteristic cyst structure is destroyed and the epithelium is non-keratinising
(Browne 1970b; Yamaguch¡ 1980).
Although mitotic figures are a normal feature of OKs, it is interesting that most of the mitotic activitiy (90%) appears in the suprabasal layers of the epithelium
(Browne 1970b; Brannon, 1977) whereas the occurrence of this activity within the suprabasal layer of the epithelium is unusual for normal stratified epithelia.
Moreover, the mitotic value of OKs reported by Browne (1975) is relatively high (3.9 to 8.0), compared to the mitotic value from other types of cysts'
2.1.4. Behaviour
OKS have clinical growth behavioural features, which are more aggresslve than those of other odontogenic cysts. Kakarantza-Angelopoulou and
Nicolatou (1990) reporled that the growth potential of epithelial cells from OKs is higher than that of the epithelium from non-keratinised dentigerous cysts. lt has also been suggested that the type of keratin present in OKs can cause differences in the behaviour of OK. ln the conclusion to their study, Brondum and Jensen (1991) pointed out that parakeratotic OKs with a thin, band-like
epithelium and palisade-like basal cells have a higher recurrence rate than
orthokeratotic OKs.
Partridge and Towers (1987) described OK cases showing paradoxical clinical
behaviour in that they featured behavioural features of both benign neoplasms
t6 and simple cysts. The authors suggested that surgeons should be aware of this because of the high rate of recurrence of the cysts.
A study conducted by Kakarantza-Angelopoulou and Nicolatou (1990) demonstrated that nine (1O.4%) of 87 OK cases recurred. The authors suggested that the cause of the recurrences were not just due to the intrinsic nature of the lesion, but was also due to the differences in surgical technique and the experience of the surgeon. A study by Crowley et al (1992) demonstrated that the recurrence rate for OKs with parakeratinised epithelium was 42.6"/o þt a49 cases), and2.2o/" for orthokeratinised OKs. There was 1 case (1 4.3"/") in which both parakeratinised and orthokeratinised features recurred. The mean recurrence time was 6.7 years, with a range from 6 months to 41 years.
The reported average time for recurrence of OKs varies between studies.
According to a study conducted by Forsell et al (1987), the recurrence rate of keratocysts in a 5-year follow-up study range from 35% to about 5O%. ln this study, the first recurrent cyst was seen 3 years after operation. The authors suggested that an active annual follow-up, over a 3-year period, should be
sufficient for practical purposes in the context of monitoring for cyst
recurrence.
t] 2.1.5. Histogenesis and pathogenesis of odontogenic keratocysts
Although it has been suggested that OKs originate from remnants of odontogenic epithelium, there is still a lack of evidence concerning the origin of OKs.
The origin of OKs has been speculated upon by Soskolne and Shear (1967).
They suggested that the cyst originates directly from dental lamina tissue, in addition to deriving from preappositional enamel organ epithelium. This suggestion was based on their examination of OKs from patients with BCNS.
They also suggested that satellite microcysts in the wall of the main cysts also originate from dental lamina related to the main cysts. Even though many studies have been carried out to search for the link between keratocysts and
BCNS, an answer has not yet been found. Some researchers have suggested, however, that there exists in some individuals a predisposition to keratocysts (Woolgar et al 1987; Lombardi et al 1995)'
Given that some researchers consider the remnants of the dental lamina to be the source of cysts, such as OKs, one would expect such cysts to be located in the gingiva or in the adjacent tissue. However, some keratocysts are
located in the ascending ramus, which does not have a relationship to a tooth
follicle or dental lamina (Shear 1992). Another suggestion concerning the
source of OKs is that they originate from basal cell offshoots or basal cell
hamartias from the overlying oral mucosa (Shear 1992).
18 There are a number of reported cases of OKs associated with BCNS. Seven (8.8%) of the 87 OK cases reported by Kakaranlza-Angelopoulou and
Nicolatou (1990) were part of the BCNS, and Farndon et al (1992) concluded that recurrent OKs are one of the characteristics of BCNS. A study conducted by Wicking et al (1994) suggested that there may be a gene alteration that is responsible for BCNS. The gene responsible for BCNS has been mapped as gq22.3-q31 in a 12-cM interval between the micro-satellite marker loci
DgS1 2.1 and D9s109 (Wicking et al 1994) or between DNA markers D9S12 and DgS53 (Farndon et al 1992). However, no study has yet been conducted to locate the gene that is responsible for OKs.
Although the histologic and behavioural features of the OKs were delineated approximately 3 decades ago, additional studies by a number of investigators have not shed any new light on the biology of these lesions' However, the relatively recent development of a variety of immunohistochemical "probes" which allow investigation of both normal and pathological tissue relationship to such things as "growth promotors" and "inhibitors" affords new opportunities for investigation of OKs.
Two such "probes" are p53 and Ki-67 and both of these can provide data
relative to the histologically expressed growth activity of tissue such as the
epithelium of OKs.
ln recent years there have been reports deriving from a small number ol
studies, which describe p53 distribution in OKs. More recently data describing
t9 Ki-67 distribution in OKs have been reported. Analysis of these data shows
that relatively small sample sizes have been employed by researchers and
that results are variable. There is a need for further investigation of p53 and
Ki-67 expression in OKs.
ln the subsequent sections of this literature review the topics of p53 and Ki-67
are reviewed with reference to OKs where relevant'
2.2. p53
p53 is recognised as the most frequently mutated gene in a wide range of
human cancers. Mutant p53 can be involved in programmed cell death,
apoptosis and cell transformation. p53 is also a protein that can regulate
transcription (Levine 1 992a&b). rI rrJ
2.2.1. lnitial identification of p53
pS3 was first identified by Lane and Crawford (1979), in studies investigating
the effects of the simian virus (SV40) on cells in culture. SV40 was found to
have the ability to transform cells in culture and to produce tumours if injected
into isogenic animals. Furthermore injected animals were found to produce
antibodies against SV4O. Subsequent analysis of animals injected with SV40-
transformed cells resulted in the discovery and partial characterisation of p53
(Levine 1992; Linzer and Levine 1979).
l 20 2.2.2. p53 Structural characteristics.
p53 consists of three components. The first part is the amino terminus, which
lies within residues 1-43. The amino acids have 5 spaced cluster domains, I
to V, which are mainly involved in p53 activity. The acid rich residues of the N-
terminus are responsible for its transcriptional activation.
The second portion is the central part of the molecule where there are highly
conserued blocks, which consist of sequence-specific DNA binding-sites,
which lie approximately within residues 100-300. Although this portion of the
p53 molecule is recognised as having the highest sequence-specific DNA
binding activity, it appears to have far less nonspecific DNA-binding activity
compared to the full-length protein. This central portion of the p53 protein is
also reported to be the location at which most mutations occur (Bargonetti et
al 1993).
The third component is the carboxyl terminus. This portion lies within
residues 300-393 where the sequences for oligomerisation and nuclear
localisation occur. This highly basic C terminus has the ability to negatively
regulate p53 binding to DNA, specifically in bacteria (Ko and Prives 1996).
This terminus is subdivided into three regions according to function. The first
region, known as a flexible linker, lies within residues 300-320 and connects
the DNA binding domain to the tetramerisation domain. The second region is
the real tetramerisation domain, which lies within res ues e IL,l
region is the extreme carboxyl terminus known as the basic rich region
þ 21 (Bargonetti et al 1993; Prives 1993; Gottlieb and Oren 1996; Ko and Prives
1 ee6).
Although the ability of p53 to regulate growth requires the full length of the protein, it appears that the carboxyl domain is the most complex and busiest portion of p53 (Prives 1993),
2.2.3. Molecular weight and location of p53 gene.
The human p53 gene is a single gene located in the shoft arm of chromosome
17, (which corresponds to chromosome 11 in mouse chromosomes)
(Benchimol et al 1985). The location is specified as position p13.1.
The p53 protein is a product of the p53 gene. Human p53 has a molecular weight of 53,000 daltons (53 kDa).
2.2.4. p53 protein levels in cells
The level of p53 is high in a wide variety of transformed cells and tumour cells,
regardless of the transforming agent (Benchimol et al 1982; Thomas et al
1983). Modulation of the hatf-life or stability of the p53 protein, is the major
mechanism for controlling the p53 level in normal and transformed cells (Oren
et al 1981; Reich et al 1983). Low levels of p53 with a half-life of about 20
minutes have been detected in normal cel reas mu
pb3 with a half-life of hours can occur in transformed cells (Reich et al 1983)
22 Researchers hold many different opinions about what induces the level of p53 in cells. One of the commonly accepted views is that p53 levels are at least paftly controlled by a post-transcriptional mechanism associated with a decline in replicative DNA synthesis. (Kastan et al 1991). Other p53 inducing agents include ionising radiation, ultraviolet light irradiation (Greenblatt et al
1994; Midgley et al 1995), or the presence of DNA strand breaks (Ko and
Prives 1996). Hypoxia, heat and starvation have also been found to induce a high level of p53 (Ko and Prives 1996).
2.2.5. Types of p53
The different behaviours of p53 are believed to be due to the different types of p53 found in various cell types. Wild type p53 and mutant p53 are the two main types of p53 involved in progression and control of the cell cycle.
2.2.5.1. Wild-type p53
Wild type p53 as found in normal cells is believed to be a negative regulator of cell growth.
Over-expression of wild type p53 in cells will block cell division in the Gr
phase of the cell cycle, which will prevent progression into the S phase (DNA
synthesis) (Levine 1992a&b). The role of p53 during the late Gr phase of the
cell cycle, acting as a transcription factor or a protein regulattng transcrlptlon
factors in the cell, can be explained in two ways. First, p53 may regulate
23 several genes or products essential for entry into the S phase of the cell cycle.
Secondly, p53 could play a role in DNA replication or the initiation of DNA replication (Lane and Benchimol 1990).
The relationship between expression of p53 and G1 arrest in cell cycles is due to the role of p53 in modulating damaged DNA before entering the S phase' p53 will initiate the arrest that allows cells time to repair and maintain genetic stability (Kastan et al 1992). This checkpoint activity of p53 does not just occur in the G1 phase, but also in the G2 phase. Transient G2 delay in human ovarian cancer cells is basically caused by an over-expression of p53.
2.2.5.2. Mutant p53
p53 mutations are the single most common genetic alteration obserued in human cancers. Mutation in tumour suppressor genes is a complex biochemical process which appears to be due to a combination of both endogenous biological processes and exogenous carcinogens'
Loss of function in alleles is thought to be one of two indicators of p53 mutation. Loss of allele function could be induced by deletions, inseftions or
re-arrangements in one or both p53 alleles. Missense mutation in one allele will produce an altered protein, while the other allele may be inactivated, be
lost or not produce gene products. The other indication for missense mutation
is increasing oncogenic activity (Levine et al 1991; Vogelstein et al 1992).
24 Mutation in tumor suppressor genes is quite distinct from that in oncogenes.
Mutation in oncogenes commonly occurs in 1 of 2 alleles of the gene, which act in a transdominant fashion out-weighing the growth suppressor effects of endogenous wild type p53 (Finlay et al 1989). Consequently, the alleles endorse mutations that fundamentally or permanently signal the cell to divide.
Mutation in oncogenes appear to arise spontaneously in somatic tissues over the lifetime of the organism and naturally occurring inherited forms of these mutations are not known.
Mutation in tumor suppressor genes occurs in several different codons, but is not randomly distributed in the gene. ln several murine and human tumours, mutation occurs in both alleles of the tumour suppressor gene (Mowat et al
1g8S; Baker et al 1989). Mostly however, mutation occurs in one allele of a tumor suppressor gene that will inactivate the function of its protein, and the second allele is lost by lessening the tumor cell or by gene conversion to homozygosity (Levine et al 1991).
Mutation of p53 has now been recognised in a wide variety of human tumours.
The origins of mutation of p53 protein are distinct in different types of tumours.
The cell type, capacity for DNA repair and metabolic differences together with
exogenous factors are among the factors that could account for differences in
the origin of p53 mutations (Hollstein et al 1991).
Temperature is one of the factors that can regulate the mutation of p53 oC, protein. Most of this protein in cells is in a mutant conformation at 39.5 and
25 oC, will be localised in the cytoplasm, whereas at 32.5 the majority of this protein is in the wildtype (normal) conformation and is located in the nucleus.
Transformed cells will multiply rapidly during the G1 phase of the cell cycle at
39.5 oC, whereas at 92.5 oC, the growth of these cells is stopped and the cells are terminated at the Gr/S border of the cell cycle. Wild-type proteins enter the nucleus at the beginning of the S phase (Hinds et al 1987; oC, Michalovitz et al 1990 and MaftinezeI al 1991). At 37 both mutant and wild type p53 proteins are present in the transformed cells and are located in the cytoplasm during the G1 phase of the cell cycle. Mutant p53 protein is not temperature-sensitive and remains in the cell cytoplasm throughout the cell cycle (Hinds et al 1987).
The description that led to the idea of p53 as an oncogene was based on studies that showed some DNA and p53 clones in culture could immortalise cells (Jenkins et al 1984; Finlay 1989; Levine 1992a&b)' The clones contributed to transformation of the cells if transfected into rat embryo fibroblasts along with the activated ras oncogene (Eliyahu et al 1984).
Transformation and immortalisation activities are thought to be two separate activities (Eliyahu et al 1985). According to Eliyahu et al (1988) and Finlay et al (1989) the wild type p53 gene does not transform cells. ln fact, as
previously described it could inhibit the transformation of cells and block the
growth of the cell in the late G1 stage of the cell cycle. On the other hand,
mutant p53 VES an oncogene a oug
26 mutant p53 protein will lead to a high potential for tumourigenic activity in cells, it will not cause gross morphological changes (Eliyahu et al 1985)'
2.2.6. Biological functions of p53
Although there are differing opinions among researchers concerning the functions of p53 (and this has led to some confusion), the functions of p53 can be categorised as falling into several main areas. For example it has been suggested that p53 could act as a checkpoint control in the cell cycle and in genetic stability (Prives and Manfredi 1993). lt has also been suggested that p53 could also have a role in programmed cell death (apoptosis) if the repair process is halted (Gottlieb and Oren 1996). A role for p53 in cellular differentiation has also been suggested at least in relation to hematopioetic lineages (Gottlieb and Oren 1996; Prives and Manfredi 1993).
To fulfill these functions, it has been suggested that p53 could act as transcription factor in gene expression in cells. Together with other proteins
(by binding to specific binding-sites), or by itself in an oligomeric protein complex, p53 could recognise DNA sequences in enhancer and/or promoter
elements of a gene such that it effectively controls that gene (Levine
1992a&b).
27 2.2.6.1. Checkpoint of cell cycle.
Checkpoint of cell cycle is a mechanism that controls and imposes dependency on each of the phases in the cell cycle. Every phase in the cell cycle should depend upon the completion of earlier events. A cell cycle with lack of checkpoint such as occurs in mutations would decrease the dependency of the events. For example mitosis could occur without completion of DNA synthesis (Hartwell and Weinert 1989).
p53 is classified as one of the tumor suppressor genes. This type of gene product negatively regulates cell growth or division. p53 has DNA binding properties and is commonly thought of as a negative regulator of cell growth. ln the cell cycle, p53 acts as a checkpoint which monitors the efficacy and completion of the obligatory events in the cycle and blocks uncontrolled cell division (Fields and Yang 1990; Haftwell and Weinert 1989; Levine 1992a&b;
Levine et al 1994).
It has also been suggested that in response to cell DNA damage, p53 activate one of two pathways in damaged cells. ln damaged cells, division will stop
until the DNA has been repaired. This pathway is known as the growth-arrest
pathway. This ability of p53 to act as a 'guardian of the genome' has been
suggested as an important role for p53 in response to DNA damage and in the
developmental processes (Clarke et al 1993; Lane 1994; Gottlieb and Oren
1 ee6)
28 2.2.6.2. Apoptosis
Wild type p53 is now believed to be also involved in programmed cell death
(apoptosis). Evidence for this role has accumulated from a number of studies.
For example, Yonish-Rouach et al (1991) showed that cells with a lack of
endogenous p53 expression, such as the murine myeloid leukaemic cell line,
will continuously lose viability and will die if the production of wild type p53 is
stimulated. lnvolvement of p53 in physiological apoptosis has been observed
in p53 deficient thymocytes. Lymphoid stem cells of the small and large
intestine in p53 (-) mice were resistant to apoptosis when the mice were
irradiated, but p53 (+) mice did not respond to the treatment (Clarke et al
1993; Lowe et al 1993).
Accumulation of p53 protein contributes to the extent of apoptosis. A clear
inverse relationship has been noted between p53 positivity and apoptosis.
Wild-type p53, which has a short half-life, is an effective apoptosis inducer,
whereas mutant p53, due to its longer halÊlife and its inability to transactivate
the bax gene (which encodes a protein that binds to bcl-2, a well-known
apoptosis suppressor) is not efficient at inducing apoptosis (Eerola et al
1ee7).
2.2.6.3. Cellular differentiation
The involvement of p53 in cell and tissue differentiation and development is
not clear. However, there is some accumulating evidence that p53 is involved
29 in such activities. For example, during normal tissue regeneration after injury, over-expression of p53 inhibits cell proliferation which will downregulate cellular growth in the healing stage. The highest expression of p53 has been recorded during the fifth day after injury, at the same time as epithelialisation of injured tissue. (Antoniades et al 1994). Expression of wild-type p53 has also been found to inhibit angiogenesis, one of the factors needed for tumour growth (Dameron et al 1994).
2.2.7. Detection of p53 in cells
p53 protein can be detected using antibodies by the standard avidin-biotin peroxidase (ABC) immunohistochemistry technique. This bridge technique is one of the most common indirect techniques used in immunohistochemistry'
The technique is carried out using a complex of primary antibodies against the antigen and avidin is bound to biotinylated peroxidase (the secondary antibody). The ABC complex is a lattice-like, three-dimensional formation and will capture several molecules of horseradish peroxidase at the site of the antigen in the section.
2.2.8. Application of p53 in tumour diagnosis.
Expression of p53 protein has been extensively used as a marker of
malignancy and pre-malignancy in humans for almost two decades. More than
S0% of human cancers have been found to express p53 protein. Detection of
pb3 mutation in cancers can contribute to increased understanding of the
30 aetiology and molecular pathogenesis of neoplasms (Hollstein et al 1991;
Levine et al 1991). Further, p53 has an important role in studies aimed at better defining the diagnosis, prognosis and treatment of a variety of cancers.
The literature relating to p53 and tumours is vast. Review of this field is
outside the reach of this study, However, brief reference is made to a small
number of recent studies to illustrate examples of work which has been
carried out.
There have been reports suggesting that colorectal cancers express
genetically altered p53 tumour suppressor genes in the range oÍ 44'8"/"
(Freedman et. al 1996) to 76"/" (llyas et al 1996) of cases. These
investigators also suggested that there is a correlation between colorectal
cancers, familial history and diet.
pS3 gene defects have been found to influence the behaviour and the
prognosis of colorectal adenocarcinomas (Benhattar et al 1996; Goh et al
1996). Benhattar et al (1996) indicated that patients whose tumours
expressed mutant p53 would have a shorter life expectancy compared to the
patients with wild type p53. These findings demonstrated that expression of
p53 protein in colorectal carcinomas could be a good indicator for diagnostic
and prognostic purposes as well as for epidemiologic assessment.
3l Overexpression p53 protein has also been used as an indicator for diagnostic, prognostic and treatment purposes for tumours in the upper aerodigestive tract. (Bennet et al 1992; Tan et al 1996; Tan and Ogden 1997.)
Expression of p53 protein has also been used as an indicator for prognostic outcome assessment in prostate cancer. Over-expression of p53 protein was reported within the range ol 5"/" - 62% of stained cells in prostate cancers
(Apakama et al 1996; Brook et al 1996). Brook et al (1996) observed that the number of positive cells and the density of staining in metastatic cases was higher compared to the stainability of p53 in primary cases (Brook et al 1996).
Katoh et al (1996) found that over-expression of p53 protein in breast cancer patients only had significant correlation with oestrogen and progesteron receptors but not with tumor size, nodal status, obesity, and histopathological tumour grade. The investigators concluded that over-expression of p53 protein was not a significant prognostic indicator for recurrence and survival in human breast cancer cases. Kristen et al (1997) suggested that p53 over- expression did not effect overall suruival of patients with node negative breast
cancer but that the expression of p53 had significantly correlated with tumour
stze.
Expression of p53 protein has also been described as a useful indicator for
prognostic purposes in relation to neoplasms such as lung cancers and
ovarian cancers. For example expression of p53 protein was found to be
32 similar in primary $2"/") and secondary (66%) ovarian cancer sites. (Daidone et al 1996).
Evidence demonstrating the presence of p53 in benign neoplasms is
equivocal. p53 expression was used by Yaziji et al (1996) to differentiate
reactive from neoplastic lesions such as in gliosis and malignant astrocytic
lesions. ln a study conducted by Lee and Teh (1994) these investigators
demonstrated that p53 expressed in all (100%) cases of pseudo-
epitheliomatous hyperplasia, 78.8"/o of cases keratoacanthoma and 75o/" ol
squamous cell carcinomas. These investigators also found that the stainability for p53 in pseudoepitheliomatous hyperplasias and
keratoacanthomas WaS less dense compared to that of Squamous cell
carcinoma.
Ueda et al (1993) found that p53 (Pb240 which stains only mutant p53) was
completely negative in all benign and reactive lesion.
2.2.9. p53 protein expression in oral lesions
Recent studies show that most invasive squamous cell carcinomas, some
carcinomas in sifu, and dysplasia from various sites in the human body,
including the oral mucosa, have higher rates of p53 expression than normal
epithelium (Ogden et al 1992b; Warnakulasurya and Johnson 1992; Regezi et
al 1995, Piffko et al 1995). Regezi et al (1995) have suggested that the
appearance of p53 in dysplastic epithelium and in situ carcinoma, and in the
33 normal appearing keratinocytes at the margin of carcinomas, may designate an early stage in the development of malignancy.
The study conducted by Regezi et al (1995) showed that 6 of 13 cases with oral epithelial dysplasia stained positively for p53 protein. The oral dysplasia of 3 of these patients progressively developed into p53 positive invasive carcinoma, 1 led into more advanced grade p53-positive dysplasia, 1 developed into pS3-negative verrucous carcinoma, and 1 represented a p53- positive dysplasia developing 5 years after treatment of a p53-positive carcinoma (Regezi et al 1995). A high level of p53 has also been reported in pre-invasive and invasive oesophageal carcinomas (Bennet et al 1992).
Ravi et al (1996) suggested that p53 protein in oral malignancies would be of a mutant p53 type. This opinion concurs with that of Kusama et al (1996) who stated that p53 gene mutation may be involved in any stage of carcinogenesis of the oral epithelium.
There is some evidence that p53 can be found in normal mucosa. ln a recent study conducted by Ogden et al (1997a&b) it was found that p53 protein was expressed on tissue taken from clinically and histopathologically normal
mucosa of cancer patients and non-cancer patients. The authors suggested that expression of p53 protein in normal mucosa did not necessarily indicate
the evidence of a further tumour.
34 The efficacy of p53 protein as a diagnostic and prognostic indicator for odontogenic cysts has yet to be confirmed. The presence of p53 has been identified in a variety of jaw cysts (Ogden et al 1 992a; Slootweg 1995; Li et al
1996). Specifically, expression of p53 positive cells has been found in the epithelial lining of keratocysts (ranging from 40 - 80% of the cysts investigated), dentigerous cysts (4O%) and radicular cysts (Ogden et al
1992a; Slootweg 1995; Lombardi et al 1995; Li et al 1996)'
2.3. Ki-67 and its functions
2.3.1 K¡-67
Ki-67 is a monoclonal antibody that reacts with a human nuclear antigen associated with cell proliferation. This antibody was first generated against antigens in proliferating cells in a Hodgkin's disease cell line (Gerdes et al
1984; Gown and Leong 1993).
Ki-67 was first identified in studies on the production of monoclonal antibodies
against nuclear antigens specific to Hodgkin's and Steinberg-Reed cells. lt
was found that most monoclonal antibodies generated against the nuclei of
the Hodgkin's cells were reactive with the nuclear structure of all the cells
tested, whereas only Ki-67 reacted with a nuclear antigen restricted to cells
that are known to proliferate (Gerdes et al 1983). Furthermore, Gerdes et al
(1984) and Gown and Leong (1993) reported that the antigen disappeared
when proliferating cells were induced to differentiate into resting cells.
35 Experiments have shown that the Ki-67 nuclear antigen is present in the S,
G2 and M phases, but is absent in the G0 phase (Gerdes et al 1983)' Ki-67 is absent in the early events of mitosis, but occurs positively in G1 after mitosis.
Despite the fact that Ki-67 has been employed in experiments to indicate proliferative activity since 1984, it was not until 1991 that Gerdes et al (1991) described the characteristics of the Ki-67 antigen through immunobiochemical and molecular biology techniques. Gerdes et al (1991) found that the Ki-67 antigen was highly susceptible to protease treatment and could not be extracted by 0.1 normal HCl, which meant that it was a non-histone protein.
Furthermore, it was found that the Ki-67 antibody recognised a huge protein doublet assembled by polypeptide chains with a molecular weight of 345 kd and 395 kd. through immunoblot analysis. The bands were detectable in lysates extracted from proliferating cells but were absent in quiescent cells
(Gerdes et al 1991).
The Ki-67 gene was described by Gerdes et al (1991) by screening a lambda gt11 human cDNA expression library and by identifing a 1095 base pair (bp)
partial clone which encoded the Ki-67 epitope. Analysis showed that the
clone consisted of three highly repetitive and very basic DNA sequences with
a high level of homology, each about 366 nucleotides in length. A very high
conserued region (62 bp long) and cysteine residue were found in the repeats
(Gerdes et al 1991 ; Sawney and Hall 1992)'
36 Fonatsch et al (1991) encoded the Ki-67 gene by using a large mRNA transcript, which was expressed in cycling cells. By using cDNA, the location of Ki-67 was then confirmed as having chromosome location10q25 (Fonatsch et al 1991).
2.3.2. MIB-1 as an alternative to Ki-67
Although Ki-67 has been widely used as a cell proliferation-associated marker since its discovery in 1984 by Gerdes et al, the use of this monoclonal antibody is restricted to frozen tissue sections only. lt is believed that the epitope of the antigen is damaged during tissue processing by such agents as fixatives. This restriction has been an impediment for retrospective studies, which mostly employ formalin-fixed and paraffin-embedded tissue. Not until
1991 was aKi-67 equivalent murine monoclonal raised. The new monoclonal antibody, called MIB-1, was generated from part of the Ki-67 molecule as
reported by Gerdes and colleagues (1991). This antibody can be used with
both formalin{ixed and routinely processed tissue.
MIB-1, MIB-2 and MIB-3 were confirmed as true Ki-67 equivalents and have
shown a similar immunostaining pattern to that of Ki-67 in frozen sections,
with the epitope reacting with MIB-1 staining more strongly than it does when
labelled with Ki-67. However, the MIB-2 antibody, is negative in paraffin
sections (Cattoretti et al 1992; Cuevas et al 1993).
-'t I The distribution of cells stained with Ki-67 and MIB-1 was repofted by
Hendricks and Wilkinson (1994) to be the same after microwave treatment of section prior to the immunostaining procedure. However, the stainability produced by the Ki-67 antibody was weaker than that of MIB-1'
Barbareshi et al (1994) found that the number of cells stained with Ki67 and
MIB-1 was slightly different in frozen and fixed sections of breast carcinoma.
They found that the mean and median values of MIB-1 Labeling Index (Ll) were almost double the values of Ki67. They concluded that the difference was possibly due to the ability of the MIB-1 epitope to sulvive in frozen and acetone fixed circumstances. These investigations also suggested that MIB-1 would be a better substitute for the Ki67 monoclonal antibody, because of its potential for use in paraffin sections in retrospective studies, and because it produces clearer staining allowing optimal use with a computer-assisted image analysis system (Barbareshi et al 1994)'
The qualities of immunostaining by cell proliferation-associated markers may be influenced by the effect of differences in fixation time and microwave heating time. Hendricks and Wilkinson (1994) reported 4 hours to be a
suitable time for fixing tonsil tissue, prior to MIB-1 immunostaining, regardless
of the duration of microwave pretreatment of sections, ln contrast the
stainability decreased if the tissue was fixed for 48 hours unless the heating
time was increased to at least 14 minutes. They also suggested that
prolonged h me up mrn id not have any erfect on
stainability. These investigators suggested that tissue should be fixed as
38 soon as possible and concluded that the effects of a postponing of fixation
may vary between tissue tYPe.
2.3.3. K¡-67/MlB-1 as indicators of cellular proliferation.
Since Ki-67 is present during the phases of the cell cycle designated as late
Gr, S, M and G2 (except for the Go phase), it has been employed for more
than a decade as a marker for cell proliferation rate as an alternative to earlier
methods of tumour proliferation, such as mitotic indices, which are restricted
to one phase of the cell cYcle.
An early repoft on the Ki-67 antigen (Gerdes et al 1984) showed that Ki-67
was absent between the Go and G1 phases of the cell cycle, but present in the rt rl,t G1 phase. The expression of the Ki-67 antigen increases as the cell cycle ''': progresses, rising during the latter half of the S phase and reaching a peak in
the G2 and M phases. Ki-67 rapidly degrades after mitosis, with the half-life of
the detectable antigen being an hour or less.
2.3.4. Expression of Ki-67 / MIB-1 in tumour cells
Cell proliferation rate is a useful indicator for determining prognosis and
choice of treatment for patients with malignancies. Levine et al (1997)
suggested that a high level of Ki-67 expression is one of few indicators for
poor prognostic and Poor suru rva rate sarcoma. a rle,
Valente et al (1994) found for oral squamous cell carcinomas that there is no
Þ 39 significant correlation between the number of Ki-67 positive cells at diagnosis
and after treatment with radiotherapy. However, these investigators found
that significant differences emerged after the first course of 10 Gy
radiotherapy.
ln recent years there have been a number of studies reporting the use of Ki67
or MIB-1 as cell proliferation markers in relation to a variety of malignant
neoplasms. The growth fraction of breast cancers can be determined by using
monoclonal antibody Ki-67. Gerdes et al (1986) demonstrated that the
number of positive cells against Ki-67 in mammary lesions varied between
benign and malignant lesions but showed no correlation between histological
grades of malignancy and size of the growth fraction in breast cancers of their
samples.
1,, rl,i
] The measurement of proliferative activity of tumour cells, as defined by MlBl
immunohistochemical analysis, has a potential prognostic value in-patient with
hepatocellular carcinomas. Ng et al (1995) found that the number of cells
stained by MlBl was significantly higher in less well differentiated, compared
to well differentiated, hepatocellular carcinomas (p=0'otz)' Although statistically insignificant, the MIB-1 score was also higher in non-
encapsulated, compared to encapsulated, hepatocellular carcinomas (Ng et al
1ees).
Zidar el n pos
epithelium of laryngeal lesions correspond to their histological grade' ln
þ 40 normal laryngeal epithelium and reactive lesions such as simple hyperplasia,
the location of most of Ki-67 positive cells is the basal layer, whereas in
abnormal and atypical hyperplasia, the number of positive cells increases and
the location of positive cells tends to be in the suprabasal layer (Hall et al
19g0; Zidar et al 1996). Similarfindings were also reported by lchikawa et al
(1997) where the number of Ki-67 positive cells increased according to the
severity of dysplastic changes in hyperplastic, precancerous and cancerous
oral lesions.
Zidar et al (1996) suggested that Ki-67 was a better proliferation marker
compared to proliferative cell nuclear antigen (PCNA). ln their study of
epithelial lesion of larynx, Ki-67 staining demonstrated a clear border of
positive stained cells of basaloid cells spinous cell layer, whereas PC-10 (an
antibody for PCNA) expressed weak positive staining cells scattering in the
spinous cell layer of laryngeal epithelium (Zidar et al 1996)'
Expression of Ki67 has been used as an indicator of the malignancy of
neoplasms. A clear relationship between MIB-1 staining and histologial
grade, tumour size, tumour type and survival rates of patients with breast
carcinomas was reported by Pinder et al (1995). The investigators also
suggested that MIB-1 labelling index can be used as a predictor for patients
suruival.
t I Li et al (1995) found that expressron was srgn rgne r ln solr
I ameloblastoma of follicular type compared to cystic tumour lining, intra luminal
þ 4l nodules and invading islands of unicystic ameloblastoma. They suggested that the results of their study indicated that epithelial proliferative potential was different in different areas of unicystic ameloblastoma and between solid and unicystic ameloblastomas. These findings concurred with the latest results of
Ong'uti et al (1997) where the follicular type of ameloblastoma had significantly higher Ll than plexiform ameloblastoma. Ong'uti et al (1997) found that there was no significant correlation between labeling index (Ll) and clinical features such as age, Sex and tumour size, but noted there was a correlation with histological pattern of the epithelium in ameloblastomas'
Katoh et al (1997) reported that Ki-67 Labeling index (Ll) of 4.7% on sections from leiomyoma of the mandible did not seem to correlate with clinical features such as rapidity of growth'
Only a very small number of studies have been carried out using Ki-67 expression to emphasize the proliferative activity of odontogenic cyst epithelium (Woolgar et al 1995; Slootweg 1996; Li et al 1996). However, those studies involving analysis of expression of Ki-67 in odontogenic cysts
had concluded that OK's have a greater number of Ki-67 positive cells
compared to other odontogenic cysts and inflammatory cysts (Slootweg 1995;
Li et al 1996).
42 CHAPTER 3
MATERIALS AND METHODS
43 3.1. Materials
3.1.1. Case selection
The odontogenic keratocysts used in this study were selected from the files dated from 1st January,1970 until 31 December, 1990, in the Depaftment of
Oral Pathology and Oral Surgery, University of Adelaide. The specimens had been received by the Department for histological assessment from dental practitioners and medical practitioners in private practice, from various clinics of the Royal Adelaide Hospital and Queen Elizabeth Hospital in Adelaide, and from the School Dental Service of South Australia.
The cases were diagnosed as OKs by an authorised pathologist of the
Department at that particular time with reference to the standard criteria characterised by Shear (1960) Brannon (1977) Browne (1970b) and the World
Health Organization (1992). Clinical data recorded in the files included the
location of the cyst, the age at which patients first noticed the appearance of the cyst, gender, relevant medical data and provisional diagnosis. The cases
selected for this study were re-examined retrospectively by the author and
one other obseruer relative to the criteria characterised by Shear (1960),
Browne (1970b), Brondum (1970), and the World Health Organization (1992).
Also documented in the files was such information as behavior and history of
e SI e, multiple, recurrent) and whether the patient had Gorlin=Goltz
44 Syndrome (BCNS). Histologically, the lining epithelium of the cyst was categorised with respect to the type of keratinisation.
Most of the OKs selected for this study were re-cut and restained. Sections five microns thick were cut and stained with hematoxylin and eosin (H&E) for routine observation and for selection of cases fqr immunohistochemical study.
From 124 OK cases recorded from 1970 to 1990, 50 cases were finally selected for immunohistochemical analysis. The selection was based on criteria such that:
1. the specimen should demonstrate the characteristic epithelial features
of OKs (WHO, 1992);
2. the OK should not be a dentigerous keratocyst; 3. the specimen should not be so severely affected by inflammation that
the characteristic features of OKs could not be clearly identified;
4. the specimen should not have undergone decalcification.
Clinical data including age and sex of the patients and site of the lesion were
recorded for these 50 cases
Nine specimens of ameloblastoma from the Oral Pathology and Oral Surgery
Department files between 1970 and 1990 were also selected for a pilot
analysis regarding expression of p53 and Ki-67'
45 Two primary antibodies were used for immunohistochemical analysis as follows
1 DO-7 - a monoclonal antibody that reacts with wild and mutant types of
the intercellular p53 protein. DO-7 is a mouse anti-human antibody,
which will show the accumulation of mutant p53 protein in tumors from
a broad spectrum of cell lineages. DO-7 is produced by DAKO
(Australia) Pty Ltd.
2 MIB-1 (Zymed) - a murine monoclonal antibody that reacts with the Ki-
67 nuclear antigen. MIB-1 reacts against a nuclear antigen expressed
in proliferating cells. lt recognizes the antigen within cells in the Gr, S,
M and G2 phases, but not within cells in the Go phase'
3.2. Methods
3.2.1. Preparation of routine histologic mater¡al
Where sufficient paraffin-embedded file tissue remained, fresh 5 micron sections were cut from blocks pertaining to the 124 cases selected for analysis. ln a number of cases insufficient file material remained and in these cases file copies of H&E stained sections were reassessed'
ln the case of freshly cut sections, 1 of 4 sections from each specimen
selected was restained using hematoxylin and eosin (H&E) and mounted on a
glass slide, p us coa -amrnop lane). Ine
46 slides were dried overnight in a 37" C oven. All fresh and previously H & E stained sections were examined to confirm the diagnosis of OK.
9.2.2.2 Assessment of clinical data and histologic mater¡al
3.2.2.1Glinical data
Clinical data relating to each case was recorded as follows
3.2.2.1.1. Age
The age of subjects was taken from the recorded clinical examination data available in the files
3.2.2.1.2. Sex
The sex of patients was recorded according to the clinical examination available in the files
3.2.2.1.3. Site
The site of occurrence of the cyst was determined on the basis of the clinical
information in the files and was recorded according to the classification
described by Raden and e 1
4',1 3.2.2.2. Histological assessment
Histological examination of the 124 original cases was carried out using an
Olympus CH optical microscope. Photomicrographs were taken using a Lei\z
Ortholux unit and Pan X film.
3.2.2.2.1. Epithelial lining
The epithelial linings of OKs in this study were classified according to the type of keratinization present. Keratinization of the epithelial lining was characterised as parakeratinised, o¡thokeratinised, or a combination of both. lrregularities, such as abnormality in thickness, rete-ridge budding and metaplasia in the epithelial lining were also recorded'
3.2.2.2.2. Number of cysts and behavior
The number of cysts occurring at one time, and the appearance of microcysts, were recorded from the history and clinical data in the files. lnformation on
recurrence was also obtained from the file or from surgery.
3.2.2.2.3. I nf lam mation
lnflammation was recorded according to its presence type and severity' The
presence and severity of inflammat WAS suc asm mo era
or severe. The presence of and the severity of inflammation in the cyst
48 capsule and lining epithelium of the cyst were criteria that were considered in selecting cases for immunohistochemical study. Only cases showing an absence of inflammation or one or two focal areas of inflammatory cells in the cyst capsule were considered for the immunohistochemical study phase.
3.2.4. lmm unohistochem ical phase
For the immunohistochemical phase of this study, the sections were dewaxed, rehydrated and placed in phosphate-buffered saline (PBS). lmmunohistochemical staining for p53 and Ki-67 (MlB-1) was performed after digestion with trypsin ll at 37"C using standard techniques.
After immunostaining, sections were examined using an optical microscope.
Stained sections were scored for the presence or absence of positive staining and according to the level of positivity of the staining for each antibody used'
Levels of positivity of staining were recorded as follows:
(-) no positive staining cells in the epithelial lining;
(+) scattered positive cells in the epithelial lining;
(++) few or moderate foci, discontinued positive cells along the epithelial
lining;
(+++) continuously positive cells in the epithelial lining'
Data deriving from this phase of study were collated and subjected to Chi-
square analysis.
Photomicrographs were taken of representative sections using a Olympus
camera and kodak print film.
49 CHAPTER 4
RESULTS
50 4.1. Clinical data
One hundred and twenty four cases of odontogenic keratocyst (OK) from 112 patients were recorded in the histopathology files Oral Pathology, Depaftment of Dentistry for the years 1970 to 1990.
4.1.1. Age distribution of total OK pool
The OK cases recorded in the Oral Pathology files between 1970 and 1990 occurred in patients between the ages of 7 and 81 years (mean 34.9 + 19)
(Figure 1). The age of 12 individuals was unknown. Although the mean age of patients involved in this study was 34.9 + 19 years, there tended to be 3 peaks of age distribution in the second, third and sixth decades (Figure 1)'
4.1.2. Age distribution of selected OK cases
Fifty cases of OK were randomly selected from the original 124. The cases
selected for immunohistochemical staining were from 46 patients with an age
range from 7 to 79 years and a mean age of 31 .6 + 1 7.4 years. The peak age
distribution of the sample was in the second decade (Figure 2). All four cases
associated with Gorlin-Goltz Syndrome occurred in patients with an age range
of age between 9 to 14 years.
5l 110
100
90
80
N o 70 I Male I Female o tr Total r 60
p a 50 t I e 40 n t s 30
20
10
0 8th 9fh T lst 2nd , 3rd 4th 5th 6th 7th
Age in decades
Figure 1: Age, sex d¡stribution of all 100 patients recorded in files as naving odontogenic keratocyst and for whom age/sex distribution data were available. T = total male and female and total male plus female. 60
50
I Female 40 Male N I o trTotal o f 30 c a s e s 20
10
0 lst 2nd 3rd 4th sth ôth 7th 8th T Age in decades
Figure 2: Age and sex distribution of the fifty OK cases selected for immunoh¡stochemical studY. T = total male and female and total male plus female' 4.1.3. Sex distribution of total OK pool subiects and immuno-
histochemistry pool subiects
Fifty-nine (52.67%) of the 112 patients were male and 53 (47'32%) were
female. One female and 3 males, were recorded as having the features of
Gorlin-Goltz Syndrome. Thirtytwo (64"/") of the 50 cysts selected for
immunohistochemical study were from female subjects and 18 (36%) were
from males (Figure 2).
4.1.4. Site distribution of total OK pool
The OKs analysed were distributed in the mandible and maxilla. Eighty-
seven (70.02%) of the 123 cases occurred in the mandible and 36 (29.08%)
of the cases occurred in the maxilla (Figure 3). Three patients had cysts in
both the maxilla and the mandible, while for 1 patient there was no record of
the site of the occurrence of the cyst.
Most of the OKs occurred in the posterior area of the jaws' Sixty-four
(7b.S6%) of the mandibular cases were in the posterior mandible and 9 cases were in the anterior mandible. The specific site of occurrence of 14
mandibular cases were not recorded'
54 140
120
El Mandible I Maxilla tr Total 100
N o 80 o f 60 G a s e s 40
20
0 R post anterior L post unknown total Site
Figure 3: Site distribution o1 123 OKs recorded in files R post - right posterior L post - left posterior Unknown - site not recorded Twenty-eighl (77.7"/") of the 36 cases occurring in the maxilla were in the posterior region, and 5 were in the anterior maxilla. The specific site of 3 cases in the maxilla was not recorded.
Of the S0 cases selected for immunohistochemical staining, 14 were in the maxilla and 36 were in the mandible. The pattern of distribution of these cases was similar to all cases archived. Eleven (78.5%) of the 14 cases appearing in maxilla were in the posterior region of the maxilla and 27 (75o/") of the 36 cases occurring in the mandible were in the posterior region of the mandible'
The specific site of occurrence of 8 cases from the mandible and one from the maxilla was not recorded (Figure 4).
56 26
@ Mandible
20 I Maxilla
N tr Total o
or5 f c a s l0 e s
5
0 R post Anterlor L post unknown Site
Figure 4: Site distribution of the fifty cases of OK selected for immunohistochemical study R post - right Posterior L post - left Posterior Unknown - site not recorded 4.2. Histologic observations
4.2.1. Epithelial lining of cases selected for immunohistochemical
analysis.
Although most most of the cysts (more than 75%) were affected by
inflammation, the cases selected for immunohistochemical staining
demonstrated the characteristic epithelial features of OKs with minimal effects
of inflammation.
The majo rity (g2%) of 50 OKs were lined by a parakeratinised stratified
squamous epithelium (Figure 5). Three (6%) cases were lined by an
ofthokeratinised stratified squamous epithelium (Figure 6). One (2"/") case
,l| _,tj had both an orthokeratinised and parakeratinised stratified squamous ',,1 epithelium as the cyst lining (Figure 7).
The thickness of the epithelial lining ranged between 3 to 8 cells. The
thickness and the arrangement of the epithelial cell lining were sometimes
disrupted by the presence of inflammation.
I 58 l- *.-Ð)e*jl .¡ '-t , , -p..,|,,3;r:È.. I L *â'¡-tg'¿ I _ -f2 ch"-' \ ,d 'rrrirt,.^__f: tû+¿ gi o* , r-tð). c 'pr0ai , , ,, ,"å ,.rã.a, J> - c {L
5 \'r
Figure 5 : Photomicrograph of an OK with a parakeratinised lining epithelium (E). Such an epithelium was present in 46 of the 50 cases of OKs examined in this study (L - cyst lumen, C - cyst capsule). Haematoxyfin and Eosin. original rnagnification X 200. Figure 6: PhotomicrograPh of lining epithelium(E). Such an e cases of OKs examined in this capsule). Haematoxylin and Eosin. original magnification X 200.
f
I Figure 7: Photomicrograph of an OK with mixed orthokeratinised and para-keratinised lining epithelium (E). Such an epithelium was present in one out of fifty cases of OKs exam¡ned in this study (L - äyst lumetr, C - cyst capsule). Haematoxylin and Eos¡n. Original magn¡fication X200.
t
I 4.2.2. lmmunohistochemical Staini ng
4.2.2.1.p53 sta¡n¡ng
Sections were examined for the presence of brownish staining nuclei (Figure
8). The degree of positivity of p53 staining varied between densely positive and faintly positive (Figure 9). Only cells classified as having unequivocal positivity were counted as positive. Cells having faint positivity (Figure 10) were scored as negative.
Twenty{wo Øa%) of the 50 OKs stained positive for p53. Most of the p53 positive stained cells were located in the supra-basal cell layers. The number of pb3 positive stained cells was higher in areas affected by inflammation.
Two of 4 cases associated with Gorlin-Goltz syndrome in this series were p53 positive whereas 2 p53 negative cases were recurrent cysts, which had been
previously en ucleated.
62 G - cyst capslue. Original magnification X 100. Figure 9: Photomicrograph of p53 staining in OK epithelium. Such positive stained nuclei (arrows) were present in 22 (Mo/o) of 50 OK cases in this study. Original magnification X 400.
I ta do ',
l¡l
Figure 10: Photomicrograph of an OK in which the epithetium was scored as negat¡ve for p53. There is general background staining of cyst lumen contents, parts of the epithelium and capsule component. some of epithelial cell nuclei (arrows) exhibit a faint brown staining. However, these were scored as negative. Original magnification X 100. 4.2.2.2. Relationship between p53 staining and inflammation
There was no relationship between p53 positivity and the presence of inflammation. Seventeen (77%) of p53 positive cysts were associated with varying degrees of inflammation, whereas 21 (75%) of pS3 negative cysts were associated with inflammation (X2 = 0.034; p < 98) (Figurel 1).
66 4.2.2.3. Ml B-1 Staining.
Forty one (92/") of the 50 OKs stained positive for this marker (Figure 12).
Most of the positive stained cells were located in supra basal cells. The intensity of positive staining varied, but was always denser and clearer than that for p53.
68 -+ B-
rlJ
'I
Figure 12: Photomicrograph of MIB-1 staining in an OK. Brownish stained nuclei indicated MIB-I positivity. Such positive stained nuclei were present in 41 l&4Yol of 50 OKs in this study. Original magnification X 200.
ì I
* 4.2.2.4. Relationship between MIB-1 positivity and inflammation
The stainability of MIB-1 in epithelial cells of OKs was analysed according to the degree of inflammation (Figure 13). Thirty-three (80.5%) of 41 MIB-1 positive specimens were associated with capsular inflammation. Five (55.5%) of 9 MIB-1 negative stained specimens demonstrated inflammation. Chi- square analysis demonstrated a poor correlation between MIB-1 positivity and the presence or absence of inflammation (X = 2.515, P< 0.2).
10 45
¡fll
r MrB-r(-) tr MrB-1 35
N 3{r o o25 f s20 a m Pl5 I e slo
5
0 None M¡Id Moderate Severe Total
lnflammation
Figure 13: Relationship between MIB-1 positive staining and severity of inflammation. 4.2.2.5. Relationship between p53 staining and MIB-1 staining
All22 p53 positive cysts were also positive for MIB-1. Of the 28 p53 negative samples 19 (67.8%) were MIB-1 positive, whereas I (32.1%) were MIB-1 negative (x' = 8.62; P < O.O1) (Figure 14). A majority of dense p53 positive nuclei appeared to correspond with the position of MIB-1 positive nuclei. All of the MIB-1 negative specimens were also negative for p53 protein. These phenomena demonstrate a possible relationship between p53 and MIB-1 stainability in the epithelium of OKs. Chi-square analysis confirmed an t association between p53 and MIB-1 positivity in the epithelium of OKs (X =
8.62, P < 0.01).
12 25
20 N o o15 r MrB-r(-) î I MIB-1 s a m 1 0 p I e s 5
0 põ3 P53 + P5:¡
Figure 14: Relationsh¡p between p53 staining and MIB-I staining. 4.2.2.6. Ameloblastoma
Nine cases of ameloblastoma were selected from the histopathology files of depaftment of Oral Pathology and Oral Surgery, Department of Dentistry. The
University of Adelaide. Table 1 shows the various subtypes of those 9 cases.
Table 1: p53 and MIB-1 stainability in ameloblastoma
Case No. Subtype P53 MIB-1
1.266177b Follicular + +
2.37172 Predominant follicular
3. 536/83 Unicystic + +
4.449185b Unicystic +
5. 127179A Basal cell type +
6.276175 Flexiform + +
7.233179 Follicular + +
8.172172 Plexiform
9. 795/85 Plexiform + +
Six (66.67%) of 9 cases of ameloblastoma in this series were positive for p53
(Figure 15) and 6 (66.67%) were positive for MIB-1 (Figure 16). However, there were more p53 positive cells compared to MIB-1 positive cells.
74 t: \ i N ri pl t %; h 4Ç t \ \l d \ .\ \ t } \ 3 Ì j* - \\ \ I +,1 I \r Ë'- \ \- - a râ \ ,t- I ) À b \ t T ls ,lrlft \ \ .ù t \ ¡ 3 ù :,,,1 '1*J ç- "jr.-
Figure 15: Photomicrograph of p53 staining in ameloblastoma. Such positive stained nuclei (arrows) were present in 6 (66.66%) of 9 ameloblastoma cases in this study. Original magnification X 400. Figure 16: Photomicrograph of MIB-I staining in ameloblastoma. Brownish stained nuclei indicated MIB-1 positivity. Such positive stained nuclei were present in 6(66.66%) of 9 ameloblastomas this study. Original magnification X 400 CHAPTER 5
DISCUSSION
17 5.1. Clinical and histological findings
The clinical findings of this present study regarding location of the cysts in the
jaws and age and sex distribution of patients are in general accordance with
previous studies (Shear 1960, Main et al 1970, Browne 1970a and 1971;
Brannon 1976, Vedtofte and Pretorius 1979, Ahlfors 1984, Brondum and
Jensen 1991). The age distribution of cases in this study is similar to that
reported by Vedtofte and Praetorius (1979), Ahlfors et al (1984) and Forsell
(1980). Similarly, the site distribution of cases in this study is concurs with the
findings of others. OKs in this study had a high predilection for location in
posterior sites of the mandible.
The microscopic features of OKs obserued in this study were similar to those
described by others. Most (92%) of the OKs in this study were lined by rl r parakeratinised stratified squamous epithelium. Three (6%) of OKs in this
study were lined with orthokeratinised stratified squamous epithelium and only 1 (2%) case had both parakeratinised and orthokeratinised stratified
squamous epithelium. This result is similar to the result of Kakaranlza'
Angelopoulou and Nicolatou (1990) and Crowley et al. (1992) where 98% and
86.2% (respectively) of their cases were lined by parakeratinised stratified
squamous epithelium.
The presence of inflammation in the cyst capsule was recorded in most
(76%) of the present cases. This finding is similar to the results of Kakaranlza-
Angelopoulou and Nicolatou (1990) where 72.4o/" of their cases were
inflamed. Whether the presence of inflammation in the cyst capsules was a
þ 18 result of previous biopsy, infection or was due to other factors could not be
determined in the present study.
5.2.Technical aspects of immunostaining
Despite the fact lhal 22 (44%) of 50 OKs in this study were positive for the
anti-mouse monoclonal antibody against p53 protein (DO-7, Dako Ltd.
Australia), which stains both mutant and wild type of p53 protein, the density
of nuclear staining varied between cases. Both faint and dense staining for
p53 antibody was seen in the epithelial cells of the 22 positive stained cases.
This obserued variation in staining density could explain some of the
differences between the results of the present study and those of other
studies as far as p53 expression is concerned as interpretation of p53
positivity versus p53 negativity could vary between different observers. :'| '.[l
The findings in this study demonstrated some similarity to and consistency
with the staining patterns reported in other studies. Ogden et al (1992a) found
that 5 (41.6%) of 12 cases of OKs were positive against CM-1 (a rabbit
polyclonal antibody raised against both mutant and wild type p53), and
Lombardi et al (1995) repofted that 15 (50%) of 30 OKs were positive against
CM-1. Slootweg (1995), on the other hand, found that 11 (84.6%) of 13 OKs
were positive against the murine monoclonal antibody BP53-12-1 (Bio-genex,
San Ramon, CA, USA) which reacts against both mutant and wild type p53.
Li et al (1996) reported that BP53-12 gave a denser and greater number of
positive-stained cells in their 12 OK samples, compared to clone 1801 and
polyclonal CM-1, which stained 38% and 71o/" of cells, respectively. ln fact,
k 79 BP53-12 positive cells were detected in the lining epithelium of all cyst types
in their samples. Lombardi et al (1995) described the intensity of staining of
p53 in OKs as weak and speckled. The differences in the results reported by
these authors could also thus be an effect of the different antibodies used.
The relationship between p53 positive cells and proliferative activity in the
epithelial lining of OKs has yet to be understood. Most of the p53 positive
cells in the present study were distributed in the suprabasal layers of the lining
epithelium. This concurs with the findings of Slootweg (1995), Li et al (1996)
and Lombardi et al (1995) who also found that most of the p53 positive cells
were located in the suprabasal layers of the lining epithelium. The location of
p53 positive cells could be related to the mitotic activity in OK lining
epithelium. Browne (1971) suggested that the majority of mitoses found in the
rt lining epithelium of OKs are not located in the basal layer, only 1Oo/" of
I mitoses occurring there.
Ogden et al (1992a) found that p53 positive cells predominantly occurred in
the basal layers. Although Ogden et al (1992a) and Lombardi et al (1995)
used the same methods and the same type of antibody, the pattern and the
intensity of staining in the study by Lombardi et al (1995) were distinctly
different. These differences remained unexplained.
The density of p53 staining in the present study was not as dense as that
I observed tor Mlts-1. P53 stained light brown whereas MIB-1 staining was I solid and dark brown showing a clear contrast, which was easily identified, as
þ 80 positive. Slootweg (1995) also described the pattern of the staining of p53 protein and Ki-67 as being different.
5.3. Pre-treatment effects on staining
Another possible explanation for the different results among studies could be the different procedures used in immunostaining. A pretreatment procedure for antigen retrieval is one of the procedures in the avidin-biotin peroxidase method that can vary among laboratories. Some laboratories prefer to use a microwave-oven pretreatment technique for antigen retrieval, while others used boiled, freshly prepared citrate/HCl buffer (10 nmol) pH 6.0 for 15 min on an electric hotplate, together with protease digestion (Slootweg 1995, Shi et al
1ee1).
Pretreatment for antigen retrieval in immunohistochemistry has been introduced to overcome the problem of the presumed effects of fixatives such as formalin, in tissues. This procedure is designed to expose antigenic sites hidden by cross-linked proteins (Gown and Leong 1993). Antigen retrieval has
been recognised as having have some advantages including the following;
better staining for long{erm formalin-fixed tissue; reduced incubation times for
primary antibodies and possible exclusion of procedures such as enzyme pre-
digestion (Shi et al 1991). ln the study, pretreatment for antigen retrieval was
carried out by microwave heating sections prior to immunostaining.
Dowell and Ogden (1996) suggested that antigen retrieval could increase
immuno-reactivity for p53 protein. Fifteen (68.18%) oÍ 22 squamous cell
81 carcinomas (SCC) and 35 (97.2%) of 36 benign lesions from their study over- expressed p53 in samples in which antigen retrieval was applied. However, only 6 (27.27o/") of 22 SCC and 1 (2.77%) of 36 benign lesions over- expressed p53 in samples in which antigen retrieval was excluded. On the other hand Li et al (1995) found that the stainability of Ki-67 in freshly cut tissue was similar to that for microwave-treated formalin-fixed tissues. No significant difference was found between the number of positive cells in sections from paraffin embedded and frozen tissue.
5.4. p53 staining
Positivity for p53 protein in any reactive lesion does not always connote mutation in the gene. Some commercially available antibodies stain both mutant and wild type p53 proteins. Accumulation of wild type p53 protein can be detected in normal tissue as a result of physiological processes, such as the maturation of haematopoietic cells (Kastan et al 1991, Dowell and Ogden
1996) and of lymphocytes with rapid proliferative activity (Mercer and Baserga
1985). ln fact Li et al (1995) suggested that there are no mutations found in exons 5 - 10 of the p53 gene of odontogenic keratocysts. This finding leads to speculation that positive cells in OKs might be caused by an accumulation
of wild type p53. Positive staining for p53 protein in OKs should therefore be
cautiously interpreted.
Moreover, over-expression of p53 protern in reactive lesions such as Kelolcls,
fasciitis, myositis, pyogenic granuloma and benign neoplasms support the
82 opinion that p53 positivity is not just the result of over-expression of mutant p53 gene but could also be the expression of wild type p53 in physiological processes (Kastan et al 1991, Dei Tos et al 1993).
5.5. MIB-1 staining
The very few previous studies of MIB-1 expression in OK lining epithelium have reported positive cells located in the suprabasal layer of the lining epithelium. Forty-one (82"/") of OKs in the present study had MIB-1 positive stained cells possibly indicating that most of the OKs in this study were in a stage of active proliferation. This result concurs with the results of other studies. ln the present study, all p53 positive cases were also positive for
MIB-1, and the pattern of the positive stained cells against both antibodies was for cells to be mainly located in the suprabasal layers. This similar pattern of site distribution of staining for both antibodies in OKs might indicate that p53 is acting as a checkpoint control in the cell cycle to either suppress and/or modulate cell proliferation.
5.6. p53 staining and inflammation
ln this study p53 expression did not seem to correlate with the presence of
inflammation. Seventeen (77%) of p53 positive cysts were associated with
varying degrees of inflammation, whereas 21 (75%) of p53 negative cysts
were associated with inflammation (X2 = 0.034; p < 98). These results may
indicate that inflammation in the capsule of OKs influences neither the lining
epithelium nor the expression of p53 protein in the lining epithelium. These
83 results do not concur with those of the study by Antoniades et al (1991), which demonstrated that p53 expression declined during active cellular proliferation in acutely-injured skin tissue and re-emerged during the healing process.
5.7. MIB-1 staining and inflammation
Even though statistically insignificant (p<0.2), the expression of Ki-67 (MlB-1) in the epithelial lining could correlate with the occurrence of inflammation in the capsule of OKs. Thirty-three (80.5%) of 41 MIB-1 positive specimens in this study were associated with capsular inflammation. Five (55.5%) of I MIB-
1 negative stained specimens demonstrated inflammation. To date there have not been any other repofts regarding the occurrence of Ki-67 antigen in inflamed odontogenic cyst capsules. The finding of a lack of correlation between MIB-1 staining and inflammation thus stands as an isolated observation at present. However, it is interesting to speculate on the lack of correlation between inflammation and p53/MlB-1 positivity. This finding would suggest that the presence of inflammation in OK cyst walls does not influence
lining epithelium proliferation rates.
5.8. Ameloblastoma
Six (66.67"/") of 9 cases of ameloblastoma in this series were positive for p53.
The results of this study were similar to those reported in other studies.
Slootweg (1995) found that 6 (66.67%) of 9 cases of ameloblastoma were
positive against p53 protein and all 9 cases of ameloblastoma were positive
for MIB-1. ln the present study, only 6 (66.67%) of 9 cases of ameloblastoma
84 were posit¡ve for MIB-1. The differences in the results of these and other studies for MIB-1 may be explained by differences in the procedures used in staining such as the type of antibody and the antigen retrieval method used.
Although the number of positive stained cells for both p53 and MIB-1 was exactly the same, a few cases showed different patterns and intensity of staining, ln contrast to the OK portion of the study, there were more p53 positive cells in ameloblastomas compared to MIB-1 positive cells. Further work aimed at analysing p53 expression in odontogenic tumours such as ameloblastoma would be of value.
85 . CONCLUSIONS
86 CONCLUSIONS
The main purposes of this study were to;
(a) assess the expression of two "growth indicator" markers (p53 and MIB- 1) in a sample of OK's larger than that previously described in the literature
(b) examine for any correlations in the expression of these markers
(c) assess the expression of p53 and MIB-1 in a pilot sample of ameloblastomas.
On the basis of the results of this study it was concluded that:
1 Approximately half (44%) of the 50 OK's examined expressed p53 positivity. This figure is within the range of values repofted by other investigators.
2 p53 and MIB-Iexpression did not seem to correlate with capsular inflammation.
3 The majorily (82%) of OK's in this study were positive for MIB-1. This observation is in general accord with the findings repofted by three other groups of investigators who have examined MIB-1 expression in OK's. lt is concluded that MIB-1 is a better marker of cellular proliferation activity in OK's than p53.
4. Although the number of cases of ameloblastomas studied was small p53 expression was more overt than MIB-1 expression.
87 .BIBLIOGRAPHY
88 a Ahlfors E, Larsson A, Sjogren S 1984. The Odontogenic keratocyst: a
benign cystic tumor?. J Oral Maxillofac Surg. 42:10-19. a Anand VK, Arrowood JP, Krolls SO 1994. Malignant potential of the
odontogenic keratocysf. Otolaryngol Head and Neck Surg. 111: 124-128. a Antoniades HN, Galanopoulos T, Neville-Golden J, Kiritsy CP, Lynch SE
1994. p53 expression during normal tissue regeneration in response to
acute cutaneous injury in swine. J Clin lnvest. 93:2206-2214. o Apakama l, RobinsonMC, Walter NM, Charlton RG, Royds JA, Fuller CE,
Neal DE, Hamdy FC 1996, bcl-2 overexpression combined with p53
protein accumulation correlates with hormone-refractoryr prostate cancer.
Br J Cancer.74'. 1258-1262. o Baker SJ, Fearon ER, Nigro JM 1989. Chromosome 17 deletions and p53
gene mutations in colorectal carcinoma. Science. 244: 217-221 . a Barbareschi M, Girlando S, Mauri FM, Forti S, Eccher C, Mauri F, Togni R,
Palma PD, Doglioni C 1994. Quantitative growth fraction evaluation with
MtBl and Ki-67 antibodies in breast carcinomas. Am J Clin Pathol. 102:
171-175.
a Bargonetti J, Mafredi JJ, Chen X, Marshak DR, Prives C 1993. A
proteolytic fragment from the central region of p53 has marked sequence-
specific DNA-binding activity when generated from wild type but not from
oncogenic mutant p53 proteln. Genes & Development. 7:2565-2574.
a Benchimol S, Lamb P, Crawford LV, Sheer D, Shows TB, Bruns GAP,
Peacock J 1985. Transformation associated p53 protein is encoded by a
gene on human chromosome 17. Somat Cell and Mol Genet. 11: 505-509.
89 Benhattar J, Cerottini JP, Saraga E, Metthez G, Givel JC 1996. P53
mutations as a possible predictor of response to chemotherapy in
metastatic colorectal carcinomas. Int J Cancer, 69: 190-192. a Bennet WP, Hollstein MC, Metcalf RA, Welsh JA, He A,Zhu S-M, Kuster l,
Resau JH, Trump BF, Lane DP, Harris CC 1992. p53 mutation and protein
accumulation during multistage human esophageal carcinogenesis.
Cancer Res. 52: 6092-6097. a Bookstein R 1997. Tumour suppressor genes in prostatic oncogenesr's. J
Cell Biochem Suppl. 19:217-223. a Brannon RB 1976. The odontogenic keratocyst. A clinicopathologic study
of 312 cases. Part l. clinical features. Oral Surg . 42: 54-72' a Brannon RB 1977. The odontogenic keratocyst. A clinicopathologic study
of 312 cases. Part Il histologic features. Oral Surg . 43: 233-255. o Brondum N, Jensen VJ 1991 . Recurrence of keratocysts and
decompression treatment, a long term follow-up of forty-four cases, Oral
Surg Oral Med Oral Pathol. 72:265-269. a Brooks JD, Bova GS, Ewing CM, Piantadosi S, Cafter BS, Robinson JC,
Epstein Jl 1996. An uncertain role for p53 gene alterations in human
prostate cancers. Cancer Res. 56: 3814-3822.
a Browne RM 1969. Ihe pathogenesis of odontogenic keratocysf. Proc lnt
Acad Oral Pathol. 28-38.
a Browne RM 1970. The odontogenic keratocyst clinical aspect. Brit Dent J.
128:225-231.
o Browne RM 197O. The odontogenic keratocyst histological features and
their correlation with clinical behaviour. Brit Dent J. 131:249-259.
90 . Browne RM 1975 . The pathogenesis of odontogenic cysts: a review. J
Oral Path. 4:31-46.
. Cattoretti G, Becker MHG, Key G, Duchrow M, Schluter C, Galle J, Gerdes
J 1992. Monoclonal antibodies against recombinant parts of the Ki-67
antigen (MlB-1 and MIB-3) detect proliferating cells in microwave-
processed formalin-fixed paraffin sections. J Pathol. 168: 357-63.
. Clarke AR, Purdie CA, Harrison DJ, Moris RG, Bird CC, Hooper ML, Wyllie
AH 1993 Thymocytes apoptosis induced by p\S-dependent and
independent pathways. Natu re 362: 849-852.
. Cohen M A, Shear M. Histological comparison of parakeratinised and
ofthokeratinised primordial cysts (keratocysfs/. J Dent S Afr. 1980 35: 161-
165
. Crawford LV, Pim DC, Bullbrook RD 1982. Detection of antibodies against
the cellular protein p53 in sera from patients with breast cancer. lnl J
Cancer. 30: 403-408.
. Crowley TE, Kaugars GE, Gunsolley JC 1992. Odontogenic keratocysts: a
clinical and histology comparison of the parakeratin and orthokeratin
variants. J Oral Maxillofac Surg. 50:22-26.
. Cuevas E, Jones DB, Wright DH 1993. Immunohistochemical detection of
tumour growth fraction (Ki-67 antigen) in formalin-fixed and routinely
processed fr'ssues. J Pathol. 169:477-478.
. Daidone MG, Benini, E, Valentinis B, Tomasic, Bolis G, Villa A, Silvetrini R
1996. p53 expression, DNA content and cell proliferation in primary and
synchronous metastatic lesion from ovarian surface epithelial-stromal
tumours. Eur J Cancer. 324: 1388-1393.
9l o Dameron KM, Volpeft OV, Tainsky MA, Bouck N 1994. The p53 tumor
suppressor gene inhibits angiogenesis by stimulating the production of
thrombospondin. Cold Spring Harb Symp Quant Biol.59:483-489. a Dayan D, Buchner A, Gorsky M, and Harel-Raviv M 1988. Ihe peripheral
odontogenic keratocysf. lnt J Oral Maxillofac Surg. 17: 81-83. o Dei Tos AP, Doglioni C, Laurino L, Barbareschi M, Fletcher CD 1993. p53
protein expression in non-neoplastic lesions and benign and malignant
neoplasms of soft tissue. Histopathology. 22: 45-50. a Domenici-Lombardo L, Amunni F, Bergamini M, Romagnoli P 1996'
Odontogenic jaw cysts: light and electron microscopical investigations.
Histol Histopathol. 1 1: 371 -377.
Dowell SP, Ogden GR 1996. The use of antigen retrieval for
immunohistochemical detection of p53 overexpression in malignant and
benign oral mucosa'. a cautionary note. J Oral Pathol Med. 25: 60-64' o Eerola A-K, Tormanen U, Rainio P, Sormunen R, Bloigu R, Vahakangas K,
Lehto V-P, Soini Y, Paakko P 1 997. Apoptosis in operated small cell lung carcinoma is inversely related to tumour necrosis and p53
immunoreactivity. J Pathol. 1 81 : 172-177 .
o Eliyahu D, Michalovits D, Oren M 1984. Overproduction of p53 antigen
makes established cells highly tumorigenic. Nature. 316: 158-160.
o Eliyahu D, Goldfinger N, Pinhasi-Kimhi O, Shaulsky G, Skurnik Y, Arai N,
Rotter V, Oren M 1988. Meth A fibrosarcoma cells express two
transforming mutant p53 species. Oncogene.3: 313-321
o Farndon PA, Del Mastro RG, Evans DGR, Kilpatrick MW 1992. Location of
gene for Gorlin syndrome. Lancet. 339: 581-582.
92 Fields S, Yang SK 1990. Presence of a potent transcription activating
sequence in the p53 protein. Science.249: 1046-1048.
Finlay CA, Hinds PW, Levine AJ 1989. Ihe proto-oncogene can act as a
suppressor of transformation. Cell. 57: 1083-1093. a Finlay CA, Hinds PW, Tan TH, Eliyahu D, Oren M, and Levine AJ 1988.
Activating mutations for transformation by p53 produce a gene product that
forms an hsc70-p53 complex with an altered half life. Mol Cell Biol. 8: 531-
539. a Fonatsch C, Duchrow M, Rieder H, Schluter C, Gerdes J 1991 .
Assignment of the human Ki-67 Gene (MKl67) to 10q25-qfer. Genomics.
11: 476-477.
O Forsell K, Forsell H, Karnberg KE 1980. Recurrence of keratocysts, a long-
term follow-up study.lnt J Oral Maxillofac Surg. 17:25-28. o Freedman AN, Michalek AM, Marshall JR, Mettlin CJ, Petrelli NJ, Black
JD, Zhang ZF, Satchidanand S, Asinruatham JE 1996. Familial and
nutritional risk factors for p53 over-expression in colorectal cancer. Cancer
Epidemiol Biomarkers Prev, 5: 285-291 .
a Gerdes J, Scwab U, Lemke H, Stein H 1983. Production of a mouthSa'
monoclonal antibody reactive with a human nuclear antigen associated
with cell proliferation. lnt. J. Cancer. 31: 13-20.
o Gerdes J, Lemke H, Baisch H, Wacker HH, Schwab U, Sten H 1984. Cell
cycle analysis of a cell proliferation-associated human nuclear antigen
defined by monoclonal antibody Ki-67. J lmmunol. 133: 1710-1715.
93 a Gerdes J, Lelle RJ, Pickarlz H, Heidenreich W, Schwarting, Kurtsiefer L,
Stauch G, Stein H 1986. Growth fractions in breast cancers determined in
situ with monoclonal antibody K¡-67. J Clin Pathol . 39: 977-980. a Gerdes J, Li L, Schlueter C, Duchrow M, Wohlenberg C, Gerlach C,
Stahmer l, Kloth S, Brandt E, Flad H-D 1991. lmmunohistochemical and
molecular biologic characterization of the cell proliferation-associated
nuclear antigen that is defined by monoclonal antibody K¡-67. Am J Pathol.
138:867-873. o Gerdes J, Becker MHG, Key G 1992. lmmunohistochemical detection of
tumour growth fraction (Ki-67 antigen) in formalin-fixed and routinely
processed fissues. (Letters to editor) J Pathol. 168:85-87. a Goh HS, Khine K, Elnatan J, Yao J, Smith, DR 1996. Molecularchanges of
colorectal cancer in Singapore. Ann Acad Med Singapore. 25: 3-10. o Gottlieb TM, Oren M 1996. p53 in growth control and neoplasia (review)
Biochim Biophys Acta. 1287 : 77 -102. o Gown and Leong, cited in Applied lmmunohistochemistry for the Surgical
Pathologist. 1993. Edited by Leong A S-Y. London, Edward Arnold.
Greenblatt MS, Bennet WP, Holstein M, Harris CC 1994. Mutations in the
p53 tumor suppressor gene: clues to cancer etiology and molecular
pathogenesr's. Cancer Res. 54: 4855-4878.
a Hall PA, Woods AL 1990. lmmunohistochemical markers of cellular
proliferation: achievements, problems and prospecfs. Cell Tissue Kinet,
23:505-522.
94 Hansen J. Keratocysts in the jaws. ln: Husted E and Hjorting-Hansen E.
(ed). Oral surgery tl (Transactions of the Second International Conference
on Oral Surgery 1967). Copenhagen, Munksgaard: pp 128-134. a Hartwell L, Weinert TA 1989. Checkpoints : Controls that ensure the order
of cell cycle events. Science. 246: 629-634. a Hauer A 1926. Ein Chlolesteatom im linken Unterkiefer unter einem
ret i n i e rte n W e i sh e its zah n. Ze itsc h r Sto m ato l. 24: 40- 49 . o Hendricks JB, Wilkinsons EJ 1994. Quality control considerations for Ki-67
detection and quantitation in paraffin-embedded tissue. J Cell Biochem
Suppl. 19: 105-1 10. o Hoffmeister B, Harle F 1985. (Cysts in the maxillofacial region a
catamnestic study on 3353 cysts). Dtsch ZahnarzllZ. 40:610-614. Cited in
Shear M, Cysts of the Oral Region 1992. Oxford Press pp 5 - 32. a Hollstein M, Sidransky D, Vogelstein B, Harris CC 1991. p53 mutation in
human cancers. Science. 253: 49-53. o lchikawa M, lshii K, Nakajima T, Mogi K 1997. The overexpression of p53
and proliferative activity in precancerous and cancerous lesions of oral
squamous epithelium. J Exp Clin Cancer Res. 16: 141-146.
lggo R, Gatter K, Bartek J, Lane D, Harris AL 1997. lncreased expression
of mutant forms of p53 oncogene in primary lung cancer. Lancet. 335:
675-679.
a llyas M, Tomlinson lP, Novelli MR, Hanby A, Bodmer WF, Talbot lC 1996'
Ctinico-pathological features and p53 expression in left-sided sporadic
colorectal cancers with and without microsatellite instability. J Pathol. 179:
370-375.
95 . Jenkins JR, Rudge K, Currie G A 1984, Cellular immortalization by a
IDNA clone encoding the transformation-associated phosphoprotein p53.
Nature.312: 651-654
. Johnson P, Chung S, Benchimol S 1993. Growth suppression of Friend
virus-transformed erythroleukemia cells by p53 protein is accompanied by
hemoglobin production and is sensitive to erythropoietin. Mol Cell Biol. 13:
1 456-1 463.
. Kakarantza-Angelopoulou E, Nicolatou O 1990. Odontogenic keratocysts:
clinicopathologic study of 87 cases. J Oral Maxillofac Surg. 48: 593-599.
. Kastan MB, Onyekwere O, Sidransky D, Vogelstein B, and Craig RW
1991 . Participation of p53 protein in the cellular response to DNA damage.
Cancer Res. 51: 6304-631 1.
. Kastan MB, Zhan Q, El-Deiry WS, Carrier F, Jacks T, Walsh WV, Plunkett
BS, Vogeltein B, Pornance AJ 1992. A mammalian cell cycle checkpoint
pathway utilizing p53 and GADD4í is defective in ataxia-telangiectasia.
Cell. 71: 587-597.
. Katoh A, Breier S, Stemmler N, Specht S, Blanock K, D'Amico F 1996. p53
protein expression in human breast carcinoma: Iack of prognostic potential
for recurrence of the disease. Anticancer Res. 16: 1301-1304.
. Ko LJ, Prives 1996 C. p53: puzzle and paradigm. Genes and Dev' 10:
1054-1072.
o Kostecka F. Ein Cholesteatom im Unterkiefer, Zeitschr Stomatol. 1929
27: 1102-1 1 08.
96 o Kunimi K, Amano T, Uchibayashi T 1996. Point mutation of the p53 gene
is an infrequent event in untreated prostate cancer. Cancer Detect Prev.
29:223-233. o Kusama K, Okutsu S, Takeda A, Himiya T, Kojima A, Kidokoro Y, Chu L,
lwanari S, Kudo l, Moro I 1996. P53 gene alterations and p53 protein in
oral epithelial dysplasia and squamous cell carcinoma. J Pathol. 178: 415-
421. a Kristen P, MullerJG, Caffier H 1997. Prognostic relevance of p53 in node
negative breast cancer. Anticancer Res. 17:2869-2871. a Lane DP, Benchimol S. p53 oncogene or anti-oncogene? Genes Dev.
1990 4: 1-8. a Lane DP, Crawford LV 1979. T antigen is bound to a host protein in SV40-
transformed cells. Nature. 278: 261 -263. a Lane DP 1993. Cancer a death in the life of p53. Nature. 362:786-788. a Leong AS-Y, Milios J 1993. An assessmenf of the efficacy of the
microwave antigen-retrieval procedure on a range of tissue antigens. App
lmmunohistochem. 1 : 267 -27 4.
a Levine AJ, Momand J, Finlay C 1991 . The p53 tumour suppressor gene
Nature, 351: 453-455.
a Levine AJ 1992. The p53 tumour suppressor gene and product. Cancer
Survey. 12: 59- 75.
o Levine AJ 1992. The p53 tumor-suppressor gene. New Eng J Med. 326:
1 350-1 351 .
97 a Levine AJ, Perry ME, Chang A, Silver A, Dittmer D, Welsh WD 1994. The
1993 Walter Hubert Lecture: the role of the p53 tumour-suppressor gene in
tumorigenesrs. Br J Cancer. 69: 409-416.
a Levine EA, Holzmayer T, Bacus S, Mechetner E, Mera R, Bolliger C,
Roninson lB, Das Gupta TK 1997. Evaluation of newer prognostic markers
for adult soft tissue sarcoma. J Clin Oncol. 15:3249-3257.
a Li T-J, Browne RM, Mathews JB 1995. Epithelial cell proliferation in
odontogenic keratocysts: a comparative immunohistochemical study of
Ki67 in simple, recurrent and basal cell naevus syndrome (BCNS)-
associated lesions. J Oral Pathol Med.24:221-226.
o Li TJ, Browne RM, Prime SS, Mathews JB 1996. p53 expression in
odontogenic keratocyst epithelium. J Oral Pathol Med. 25: 245-55.
o Linzer Dl, Levine AJ 1979. Characterization of a 54 K dalton cellular SV40 ,t rJ tumor antigen present in SV4)lransformed cells and uninfected '' embryonal carcinoma cells. Cell. 17:43-52.
a Lombardi T, Odell EW, Morgan PR 1995. P53 immunohistochemistry of
odontogenic keratocysts in relation to recurrence, basal-cell budding and
basal-cell naevus. Archs Oral Biol.40: 1081-1084.
a Lowe S, Schmitt EM, Smith S, Osborne BA, Jacks T 1993. p53 is required
for radiation induced apoptosis in mouse thymocytes. Nature. 362: 847-
849.
o Lund VJ 1985. Odontogenic keratocyst of the maxilla: a case report. Br J
Oral Maxillofac Surg. 23: 210-215.
a Main DM 1985. Epithelial jaw cysts: 10 years of the WHO classification. J
Oral Pathol. 14: 1-7.
q 98 o Martinez J, Georgoff l, Martinez J, Levine AJ 1991 . Cellular localization
and cell cycle regulation by a temperature-sensitive p53 protern. Genes &
Dev.5:151-159.
Mercer WE, Baserga R 1985. Expression of the p53 protein during the cell
cycle of human peripheral blood lymphocytes. Exp Cell Res. 160: 31-46.
a Michalovitz D, Halevy O, Oren M 1990. Conditional inhibition of
transformation and of cell proliferation by a temperature-sensitive mutant
of p53. Cell. 62: 671-680.
a Midgley CA, Owens B, Briscoe CV, Thomas D B, Lane DP, Hall PA 1995.
Coupling between gamma irradiation, p53 induction and the apoptotic
response depends upon cell type in vivo. J Cell Sci. 108: 1843-1848.
a Mikulicz J 1876. Beitrag zur Genese der Dermoide am Kopfe. Weiner
Med Wochenschrift. 26: 953-956, 983-986, 1004-1008.
:ll '1..! o Moos KF, Rennie JS 1987. Squamous cell carcinoma arising in : mandibular keratocyst in a patient with Gorlin's syndrome. British J Oral
Maxillofac Surg. 25: 280-284.
a Mowat M, Cheng A, Kimura N, Bernstein A, Benchimol S 1985.
Rearrange- ments of the cellular p53 gene in erythroleukaemia cells
transformed by Friend virus. Nature. 314: 633-636.
Murray AW 1992. Creative blocks: cell-cycle checkpoints and feedback
controls. Nature. 359: 599-604.
a Ng lOL, NaJN, Lai ECS, Fan ST, Ng M 1995. Ki-6Tantigen expression in
hepatocellular carcinoma using monoclonal antibody MlBl. Am J Clin
I I Pathol. 104: 313-318.
Þ 99 a Nigro JM, Baker S, Preisinger AC, Jessup JM, Hostetter R, Cleary K,
Bigner S, Davidson N, Baylin S, Devilee P, Glover T, Collins FS, Weston
A, Modali R, Harris CC, Vogelstein 81989. Mutations in the p53 gene
occur in diverse human tumour types. Nature. 342:705-707.
o Ogden GR, Chisholm DM, Kiddie RA, Lane DP 1992. p53 protein in
odontogenic cysts: increased expression in some odontogenic keratocysts.
J Clin Pathol. 42:1007-1010.
a Ogden GR, Kiddie RA, Lunny DP, Lane DP 1992. Assessment of p53
protein expression in normal, benign, and malignant oral mucosa. J Pathol.
166: 389-394.
a Ogden GR, Chisholm DM, Morris AM, Stevenson JH 1997.
Overexpresslon of p53 in normal oral mucosa of oral cancer patients does
not necessarily predict further malignant disease. J Pathol. 182: 180-184.
Ogden GR, Hall PA 1997, Field change, clonality, and early epithelial
cancer: possible lesions from p53. J Pathol. 181: 127-129. (Editorial.)
a Ong'uti MN, Cruchley AT, Howells GL, Williams DM 1997. K¡-67 antigen in
ameloblastomas: correlation with clinical and histological parameters in 54
cases from Kenya. lnt J Oral Maxillofac Surg.26:376-379.
Oren M, Maltzman W, Levine AJ 1981 . Post-translational regulation of the
54K cellular tumor antigen in normal and transformed cells. Mol Cell Biol.
1 : 101-1 10
a Partridge M, Towers JF 1987. The primordial cyst (odontogenic keratocyst)
: its tumor-like characteristics and behaviour. Br J Oral Maxillofac Surg. 25: I I 271-279.
,
h 100 OF
o Philipsen HP 1956. Om keratocyster (kolesleatom) i kaeberne. RA
Tandlaaegebladet. 60: 963-981 . a Piffko J, Bankfalvi A, Ofner D, Berens A, Tkotz T, Joos U, Bocker W,
Schmid KW 1995. Expression of p53 protein in oral squamous cell
carcinomas and adjacent non-tumourous mucosa of the floor of the mouth:
an archival immunohistochemical study using wet autoclave retrieval. J
Oral Pathol Med. 24:337-342. a Pindborg JJ, Hansen J 1963. Studies on odontogenic cyst epithelium. 2.
Clinical and roentgenologic aspects of odontogenic keratocysts. Acta
Pathol Microbiol Scand. 58: 283-294.
a Prives C 1993. Doing the right thing: feedback control and p53. Curr Opin
Cell Biol. 5:214.218. Review.
a Prives C, Manfredi JJ 1993. The p53 tumor suppressor protein: meeting
review. Genes Dev. 7: 529-534.
a Rachanis CC 1978, Shear M. Age-standardised incidence rates of
primordial cyst (keratocyst) on the Witwatestrand. Community Dent Oral
Epidemiol. 6: 296-299.
o Ravi D, Nalinakumari KR, Rajaram RS, Nair MK, Pillai MR 1996.
Expression of programmed cell death regulatory p53 and bcl-2 proteins in
oral lesions. Cancer Lett.. 105: 139-146.
Raycroft L, Wu H, Lozano G 1990. Transcriptional activation by wildlype
but not transforming mutants of the p53 anti-oncogene. Science. 249:
1 049-1 051.
l0l a Regezi JA, Zarbo RJ, Regev E, Pisanty S, Silverman S, Gazil D 1995. p53
protein expression in sequential biopsies of oral dysplasias and in situ
carcinoma. J Oral Pathol Med.24: 18-22. a Reich NC, Oren M, Levine AJ 1983. Two distinct mechanisms regulate
the levels of a cellular tumor antigen, p53. Mol Cell Biol. 3:2143-2150. a Reiter RE, Anglard P, Liu S, Gnarra JR, Lineham WM 1993. Chromosome
17p deletions and p53 mutation in renal cell carcinoma. Cancer Res. 53:
3092-3097. o Robinson HBG 1945. Classification of cysts of the jaws. Am J Orthod
Oral Surg.. 31: 370-375.
Sawhney N, Hall PA 1992. Ki-67 - structure, function, and new antibodies.
J Pathol. 168: 161-162. a Scully C, Burkhart A 1993. Iissue markers of potentially malignant human
oral epithelial lesions. J Oral Pathol Med.22:246-256. a Shear M 1992. Cysts of the Oral Region. Oxford Press. pp 5 - 32.
Shear M 1960. Primordial cyst. J Dent Ass. 15:211-217. a Shi SR, Key ME, Kalra KL 1991. Antigen retrieval in formalinlixed, paraffin-embedded tissue: an enhancement method for
immunohistochemical staining based on microwave oven heating of tissue
sections. J Histochem Cytochem.39: 741-748.
o Slavin J 1996. The role of cytokines in wound healing. J Pathol. 178: 5-10.
Slootweg PJ 1995. p53 protein and Ki-67 reactivity in epithelial
odontogenic lesions. an immunohistochemical study. J Oral Pathol Med.
24:393-397
102 a Soskolne WA, Shear M 1967. Obseruations on the pathoenesis of
primordial cysts. Br Dent J.123:321-326. a Soussi T, Caron de Fromentel C, Mechali M, May T and Kress M 1987.
Cloning and characterization of a aDNA from Xenophus laevis coding for a
protein homologusto human and murine. Oncogene.l:71-78. a Tan LK, Dowell SP, Ogden GR 1996. Antigen retrieval: p53 staining in
benign, pre-malignant and malignant tissue of the larynx. Clin Otolaryngol.
21:147-150. a Tan LK, Ogden GR 1997. P53 over-expression in laryngeal carcinoma is
not predictive of response to radiotherapy. Oral Oncol. 33: 177-181. a Thomas R, Kaplan L, Reich N, Lane DP, Levine AJ 1983.
Characterization of human p53 antigens employing primate specific
monoclonal antibodies. Virology. 131 : 502-517. a Toller P 1967. Origin of growth of cysts of the jaws. Ann R Coll Surg Engl.
40: 306-336.
a Ueda Y, Dockhorn-DworniczakB, Blasius S, Mellin W, Wuisman P, Bocker
W, Roessner A 1993. Analysis of mutant p53 protein in osteosarcomas
and other malignant and benign lesions of the bone. J Cancer Res Clin
Oncol. 119: 172-178.
a Valente G, Orecchia R, Gandolfo S, Arnaudo M, Ragona R, Kerim S,
Palestro G 1973. Can Ki67 immunostaining predict response to
radiotherapy in oral squamous cell carcinoma? J Clin Pathol. 47:109-112.
a Vedtofte P, Praetorius F 1979. Recurrence of the odontogenic keratocyst
in relation to clinical and histological features. A 20 year follow-up study of
72 patients. lnt J Oral Surg. 8: 412-420.
103 o Vogelstein B, Kinzer KW 1992. p53 function and dysfunction. Cell. 70:
523-526. a Voorsmit RA, Stoelinga PJ, van Haelst UJ 1981. The management of
keratocysfs. J Maxillofac Surg. 9:228-236. o Wang Y, Reed M, Wang P, Stenger JE, Mayr G, Anderson ME, Schwedes
JF, Tegtmeyer P 1993. p53 domains: identification and characterization of
two autonomous DNA-binding regions. Genes & Development. 7: 2575-
2586. o Warnakulasurya KAAS, Johnson NW 1992. Expression of p53 mutant
nuclear phosphoprotein in oral carcinoma and potentially malignant oral
lesions. J Oral Pathol Med. 21:404-408. o Wicking C, Berkman J, Wainwright B, Chenevic-Trench G 1994. Fine
genetic mapping of the gene for nevoid basal cell carcinoma syndrome.
Genomics. 22: 505-511.
Wilson DF, Shear M. Developmental and inflammatory cysts of the iaws.
ln Prabhu SR, Wilson DF, Daftary DK, Johnson NW, (Eds.) Oral Diseases
in the Tropics. 1992 Oxford University Press. pp 687 - 693. a Wood MW, Medina JE, Thompson GC, Houck JR, Min K-W 1994.
Accumulation of the p53 tumour-suppressor gene product in oral
leukoplakia. Otolaryngol Head Neck Surg. 111: 758-763. a Woolgar JA, Rippin JW, Browne RM 1987. A Comparative histological
study of odontogenic keratotysts in basal cell naevus syndrome and
control patients. J Oral Pathol. 16: 75-80. a Yan JJ, Tzeng CC, Jin YT 1996. Overexpresslon of p53 protein in
squamous cell carcinoma of buccal mucosa and tongue in Taiwan: an
704 immunohistochemical and clinicopathological study. J Oral Pathol Med. 25:
55-59.
Yaziji H, Massarani-Wahai R, Gujrati M, Kuhns JG, Martin AW, Parker JC
Jr 1996. Role of p53 immunohistochemistry in differentiating reactive
gliosis from malignant astrocytic lesions. Am J Surg Pathol 20: 1086-
1 090. a Yonish-Rouach E, Resmitzky D, Lotem S, Sachs L, Kimchi A, Oren M
1991. W ildiype p53 induces apoptosis of myeloid leukaemic cells that is
inhibited by interleukin 6. Nature. 352:345-347. a Zidar N, Gale N, Cor A, Kambic V 1996. Expression of Ki-67 antigen and
proliferative cell nuclear antigen in benign and malignant epithelial lesions
of the larynx. J Laryngol and Otol. 110:440-445.
105